Time-value curves to provide dynamic QoS for time sensitive file transfers
0 likes250 views
This patent application describes a method and apparatus that utilizes time-value curves to manage quality of service in dynamic resource allocation for time-sensitive file transfers. By generating time-value curves for tasks, the invention optimizes mission performance by managing resource allocation, particularly for communication bandwidth. The approach allows for flexible resource scheduling and prioritization based on changing task demands in a workflow.
![US 2008/0040630 A1
TIME-VALUE CURVES TO PROVIDE DYNAMIC
QOS FOR TIME SENSITIVE FILE TRANSFERS
RELATED APPLICATIONS
[0001] This application claims priority under 37 CPR.
§l.ll9(e) to provisional patent application 60/721,757 ?led
Sep. 29, 2005 and incorporated herein by reference.
FIELD OF THE INVENTION
[0002] This invention relates generally to resource man
agement and more particularly to a method and apparatus for
generating and using time-value curves for resource man
agement.
BACKGROUND OF THE INVENTION
[0003] The effectiveness of any mission is heavily reliant
upon the ability of an underlying infrastructure to respond to
the dynamic requirements of the mission. Typically missions
are layered upon an existing resource infrastructure such that
the mission becomes merely a set of tasks that is supported
by the infrastructure. Layering a mission upon an existing
infrastructure typically mis-utiliZes key resources and
increases the di?iculty in detecting performance degradation
or partial failures that adversely affect the mission. Allocat
ing speci?c resources to a mission is technically challenging
and error prone. It Would be desirable to identify a mission
architecture Which Would overcome the problems of the
prior art.
SUMMARY OF THE INVENTION
[0004] According to one aspect of the invention a method
of allocating a resource to a plurality of resource consumers
includes the step of generating a time-value curve de?ning
a temporally dynamic priority of a resource consumer over
a time period and using the time-value curve to allocate the
resource to the consumer over the time period. In one
embodiment, the resource is communication bandWidth, and
the resource consumers include one or more tasks in a
mission Work?oW. Such an arrangement permits the sched
uling, pre-empting and trade-off of bandWidth betWeen
different mission tasks to optimiZe deployment of mission
tasks and concomitantly optimiZe mission performance.
[0005] According to another aspect of the invention, a
method of allocating a resource to a plurality of resource
consumers includes the steps of generating a time value
curve for each one of a plurality of tasks in a Work?oW, the
time-value curve de?ning a temporally dynamic quality of
service to be provided to the task.
[0006] According to a further aspect of the invention, a
policy engine comprises generic task pro?le information and
time value curve generation logic operably coupled to
receive task speci?c information and to calculate a time
value curve for the task using the generic task pro?le
information and the task speci?c information.
[0007] According to a further aspect of the invention, a
device comprising a computer readable medium accessible
by a resource allocator is described. The computer readable
medium stores a data structure comprising, for each one of
a plurality of enqueued tasks associated With the resource, a
time value curve de?ning a temporal priority to provide to
the associated task, the data structure being accessed by the
Feb. 14, 2008
resource controller to control access to the resource. These
and other aspects of the invention Will be described With
regard to the attached ?gures.
BRIEF DESCRIPTION OF THE FIGURES
[0008] FIG. 1 is a block diagram of a mission architecture
in Which the present invention may be implemented;
[0009] FIG. 2 is a block diagram illustrating several com
ponents that may be included in a policy engine to generate
time-value curves of the present invention;
[0010] FIGS. 3A-3H are diagrams illustrating different
types of time-value curves that may used as a basis for
time-curve generation by a policy engine of the present
invention;
[0011] FIG. 4 is a How diagram illustrating exemplary
steps that may be performed by the policy engine of FIG. 2
to generate time-value curves for one or more tasks in a
Work?oW;
[0012] FIG. 5 is a diagram of an exemplary netWork or
service resource manager Which may use the time value
curves of the present invention to control allocation of
resources to Work?oW tasks;
[0013] FIG. 6 is a timing diagram illustrating several
exemplary time-value curves that may be assigned to tasks
in a Work?oW;
[0014] FIGS. 7A and 7B are timing diagrams provided to
illustrate the addition of time value curves as tasks are added
to a Work?oW;
[0015] FIGS. 8A and 8B are used to describe an embodi
ment of the present invention supporting a surveillance
Work?oW;
[0016] FIG. 9 is a processing timing diagram Which pic
torially illustrates the bandWidth smoothing capabilities of
time-value QoS servicing of a Work?oW;
[0017] FIG. 10 is a pictorial representation of task selec
tion using time-value curves in the example of FIG. 9;
[0018] FIGS. 11A and 11B are timing diagrams used to
illustrate a stalled task drop capability of the present inven
tion;
[0019] FIG. 12 is a How diagram illustrating exemplary
steps that may be performed by a policy engine performing
task dropping such as that illustrated in FIGS. 11A and 11B;
[0020] FIG. 13 is a graph illustrating the performance of
TV QoS versus Fixed QoS as the offered load increases; and
[0021] FIG. 14 is a graph illustrating hoW TV QoS pro
vides for graceful netWork failure.
DETAILED DESCRIPTION
[0022] FIG. 1 illustrates a mission architecture 1 in Which
the present invention may advantageously be used to opti
miZe mission resource utiliZation. In one embodiment the
mission architecture 1 is an adaptive netWork that uses
Work?oW Locked Loops (WLL) for mission resource selec
tion and access management. The embodiment of FIG. 1
includes a Mission Resource Manager (MRM) 10 and a
SoftWare Controlled NetWork (SCN) 12. The MRM 10
essentially orchestrates the use of mission resources by a](https://image.slidesharecdn.com/us20080040630-141031124919-conversion-gate02/85/Time-value-curves-to-provide-dynamic-QoS-for-time-sensitive-file-transfers-16-320.jpg)
![US 2008/0040630 A1
mission by identifying mission tasks and allocating mission
resources to the mission tasks. Although the MRM may be
described beloW as including certain components, it should
be understood that an MRM is not a required to implement
the present invention. For example, a set of Mission Goals
statements may be input into the Policy Engine to help form
the Time-Value curves via an administrator at a GUI or
automated via a machine interface or even a Web Services
Interface.
[0023] An exemplary MRM 10 includes a Goal to Service
translation Unit 12. The Goal to Service translation Unit
may be a semantic tool Which maps a high level goal
de?nition, such as a natural language de?nition, into real
language mission goals into policy statements. One exem
plary method for performing Goal to Service translation is
described in Ser. No. (Attorney Docket No.: 120
505), entitled Mission Goal Statement to Policy Statement
Translation, by Travostino et al, ?led , incorporated
herein by reference.
[0024] Policy statements are translated into Work?oWs
using the Goal to Service Translation Unit (GSTU) 12, the
Constraint Analyzer (CA) 15, the Discovery Engine and
Scheduler (DES) 13 and the Directory 16. Together these
components identify, quantify and qualify services that are
reserved for a mission as described in patent application Ser.
No. (Attorney Docket No. 120-506) for Work?oW
Locked Loops to Enable Adaptive NetWorks, by Travostino
et al, ?led .
[0025] For example in one embodiment the directory 16
identi?es netWork resources and services that may be allo
cated to missions. The DES 13 identi?es resources and
services that are available for use as an infrastructure of the
mission. The CA 15 evaluates the infrastructure identi?ed by
the DES 13 in vieW of mission constraints. The CA 15 and
DES 13 Work in concert With the goal translator to identify
the mission infrastructure. An Engine uses the infrastructure
and mission information to generate mission Work?oWs,
Which are forWarded to the SCN 14.
[0026] The SCN 14 generally controls the execution of the
mission and the use of the mission resources during execu
tion of the mission. The SCN 14 includes an Execution
Manager (EM) 22, a Policy Engine (PE) 24, a NetWork
Services Manager (26), a NetWork Resource Manager
(NRM) 28 and a NetWork Resource and Topology Discovery
resource 27. Together these components orchestrate the
execution of the mission Work?oW using the identi?ed
mission resources.
[0027] For the purpose of this application a mission Work
How is a collection of tasks that are performed by a collec
tion of actors, Wherein an actor may be any entity that uses
a mission dedicated resource. A mission dedicated resource
is any resource that is at least partially dedicated to a
mission.
[0028] Mission tasks are dynamic, and therefore during
the course of the mission different tasks are initiated and
terminated. At any given time during the mission, mission
tasks may seek access to a shared mission resource. Accord
ing to one aspect of the invention, each task has an associ
ated time-value curve indicating a temporal priority of the
task over the task lifetime. The time-value curves are used
during resource allocation to intelligently allocate resources
Feb. 14, 2008
to tasks to optimiZe the performance of the mission. Each
point in a time-value curve identi?es a priority of QoS to be
associated With the task for the point in time. This alloWs
QoS to be adjusted in real time based on conditions in the
?eld, thereby adding the degree of temporal granularity to
the QoS structure. Controlling temporal granularity of QoS
helps resolve the con?icts of congestion and While supplying
all the resources to the mission Within the mission time
frame, Without having to expand the siZe of the netWork.
[0029] In one embodiment the PE 24 generates a time
value curve for each mission task. The time value curves are
forWarded to the EM 22. At resource allocation intervals the
EM references the relative values of the time-value curve of
each tasks that are active at that time instance to retrieve a
time-value for each task, Wherein an active task is any task
that has been initiated and not terminated. Using any one of
a variety of selection methods, such as highest value, highest
Weighted value, etc., the time values are compared to select
one of the active tasks.
[0030] FIG. 2 illustrates several exemplary components
that may be included in a PE to generate time-value curves
for tasks. The PE 24 includes a Work?oW parser 31, a generic
task pro?le store 34, and time-value (TV) curve generation
logic 32. The Work?oW parser receives Work?oW tasks from
the Execution Machine (EM). The task may have been
initiated in response to a mission trigger; for example, in a
surveillance mission, a mission trigger may be the capturing
of an image at a sensor point in a surveillance area. The tasks
may be triggered by partial or thorough completion of
another task. HoWever the task is triggered, it is forWarded
to the Work?oW parser 31 Which extracts task information
that may be used to build the time-value curve. Any obtain
able information Which may affect a priority or quality of
service to provide to the task may be provided to the
time-value curve generation logic, including, for example,
information regarding an actor of the task, a location of the
task, a security level of the task, constraints of the tasks,
dependencies of the task, poWer requirements of the task,
execution time for the task, etc. Information regarding
services and netWork elements that are associated With the
task may also be used When calculating the time-value
curve. The present invention is in no Way limited to the use
of any particular task information in generating a time-value
curve for priority scheduling of the task.
[0031] The generic task pro?le store 34 may be used to
store pre-determined task pro?les. The predetermined task
pro?les are tailored to the resource requirements of various
pre-de?ned tasks. For example, a surveillance task that
performs high resolution image capture Would have a time
value curve that temporally aligns the resource priority With
the resource usage characteristics of the task. The resource
requirements of a task may be selected from a set of
pre-de?ned pro?les, or obtained by monitoring and pro?ling
resource usage in advance to generate a time-value pro?le
for the task. FIGS. 3A-3H illustrates several representative
time value curve pro?les. A time value-curve pro?le such as
that of FIGS. 3A and 3B may be applied to a speci?c task
identi?ed as having a relatively ?at QoS. Pro?les 3C and 3D
may be used to apply a linear QoS to a task. Impulse pro?les
3E and 3E may be used to provide high priority to a task for
a short time. Time value curves of 7G and 7H illustrate
curves that are constructed for a particular QoS pro?le.](https://image.slidesharecdn.com/us20080040630-141031124919-conversion-gate02/85/Time-value-curves-to-provide-dynamic-QoS-for-time-sensitive-file-transfers-17-320.jpg)
![US 2008/0040630 A1
[0032] FIG. 4 illustrates an exemplary process that may be
performed by the PE 24 to construct a time-value curve for
a task. At step 42 a generic pro?le associated With the task
is retrieved from the pro?le store 34. At step 44, the
time-value curve is piece-Wise constructed using the base
pro?le in conjunction With task speci?c variables, con
straints and resource information using techniques knoWn to
those in the art. At step 46 the time-value curve is stored With
the task. The TV augmented task is then forWarded to the
EM 22.
[0033] FIG. 5 illustrates exemplary components that may
be included in an Execution Manager 22 that uses the
time-value curves of the invention. The EM 22 Would
generally include task queues 52 for storing active mission
tasks. The task queues may be organiZed by mission
resource. A resource controller 56 may include or otherWise
have access to the Time-Value curves for the tasks, for
example in a time-value queue associated With the task
queue. At each mission resource allocation instance the
resource controller examines the time-values of each active
task associated With a resource at that instance, and selects
one of the tasks based on the time-value curve.
[0034] FIG. 6 is a timing diagram illustrating illustrative
time value curves for 3 tasks A (62), B (64) and C (66). The
generic task pro?le for each of the curves is shoWn by beloW
Equation I:
P(T)=Min(mT+b, MaxP) Equation I
[0035] Where P is the priority, or quality of service,
associated With the task at time T, m is derived based on the
value of the task, the lifetime of the task and the execution
time of the task and b is determined according to a value of
the task. The ?at portion on each of the time-value curves in
this embodiment represents the time needed to complete the
task. Max (P) represents the highest priority to allocate to the
task.
[0036] Thus in FIG. 6 time value curve C is a loWer value
task, Which takes a relatively short period of time to com
plete. Curve B has the highest initial priority task, With a
shorter lifetime than curve C. Curve A has the second
highest initial priority, but the shortest lifetime. Using strict
priority allocation technique, task B Will obtain access to the
resource until time T1. At time T1, access begins to be
granted to task A, With task A reaching its maximum priority
at time 2 to alloW it to be at maximum priority for the time
period associated With executing the task. The time-value
curves may thus be used provide a temporal priority of the
task Which ensures that the task can complete in its lifetime.
[0037] FIGS. 7A and 7B are pictorial representations of a
mission Work?oW, provided to illustrate the dynamic nature
of Work?oW tasks; tasks are continually instantiated and
terminated during the Work?oW. For example, at time T4, a
loW priority task 68 is triggered. At time T7 a higher priority
task is triggered. At any resource allocation period, the time
value curves may be used to real-time, dynamic resource
management for a dynamic mission Work?oW.
[0038] FIGS. 8A and 8B Will be used to describe an
exemplary embodiment of the present invention for use in a
particular mission. The mission of FIG. 8 is a surveillance
mission Which seeks to identify individuals Within a de?ned
radius. The mission comprises dependent tasks, With a ?rst
set of tasks being associated With the detection and collec
Feb. 14, 2008
tion of information for each an individual that comes Within
a radius, a second set of tasks being associated With col
lecting high resolution information regarding individuals
Within the radius, a third set of tasks being associated With
collecting increasingly images at very high resolution. Each
of the tasks of in the mission use netWork bandWidth for the
exchange of information associated With the target. The
individuals are grouped into loW value targets (initial prior
ity 0); mid value targets (initial priority value 1) and high
value targets (initial priority value 2). The targets are
detected using mission resource sensors distributed Within
the radius. A detection of the target results in an attempt to
collect information about the target to determine if the target
is a target of interest. The collection of information may
include the transmission of identi?cation information, such
as a loW resolution or high resolution image. The type of
collection to be performed for each target is a task in the
above described architecture. The task is assigned a time
value curve Which provides a temporal priority to the
collection of information for the particular target.
[0039] The time value curves of FIG. 8B are associated
With tasks to be performed on three targets. A ?rst task 72 is
associated With a mid priority target Which has been sensed
in area DS3. There are only 3 minutes to collect information
regarding the target before the target leaves the area. Thus
the time-value curve is adapted for this target to ensure that
data collection for this target can occur Within the target
tracking task lifetime. A second task 74 is associated With a
target in area DSa. The target has been identi?ed as a high
value target, perhaps as a result of a loW-resolution image
mapping task. The target is exiting the area in 4 minutes. The
slope and intercept of the time-value curve re?ect the
perceived target value and detected target location, and
ensures that adequate processing of the target may be
completed before the target is lost. Time value curve 76 is
associated With a task generated in response to target detec
tion in area DS2b. There are 5 minutes until the target quits
the area, and thus time-value curve 76 re?ects the loWer
quality of service level to be provided to the collection of
information regarding this target.
[0040] FIG. 9 is a data How diagram intuitively illustrating
the bene?ts of using the time-value curves of the present
invention to provide real-time temporally prioritized
resource allocation in to mission tasks. As Will be seen from
the beloW description, the time-value curves e?fectively
smooth and optimiZe resource allocation among the tasks.
Optimizing the resource allocation improves the completion
rate of mission tasks and increases the chances of mission
success.
[0041] Data series 80 illustrates a stream of resource
requests that are received in a data stream. The performance
of the mission depends upon being able to process each task
in the data stream during its lifetime; if tasks are dropped or
delayed, mission information is lost. In FIG. 9, each data slot
represents a one minute interval, and assume that the Burst
tra?ic received in interval 1 includes high, medium and loW
priority tasks having respective lifetimes of three, four and
?ve minutes.
[0042] Data How 90 illustrates an exemplary result of
resource allocation for a mission using Fixed QoS resource
allocation methods. In a typical Fixed QoS resource alloca
tion method, a ?xed bandWidth percentage is allocated to](https://image.slidesharecdn.com/us20080040630-141031124919-conversion-gate02/85/Time-value-curves-to-provide-dynamic-QoS-for-time-sensitive-file-transfers-18-320.jpg)
![US 2008/0040630 A1
each task. During interval 1 of Data ?ow 90, Burst Tra?ic for
three tasks is received. The prior art resource allocator,
receiving no information from the ?eld regarding a temporal
priority of the tasks, uses a Fixed QoS to allocate resources
to the tasks. Each of the three tasks is provided one third of
the bandwidth. At the end of interval 1 (or one minute into
the task), each of the three pending tasks in group A is 1/3
complete. During interval 2, an additional task group B is
received which includes two additional tasks. Assume group
B includes a medium and high priority task with respective
lifetimes of four and ?ve minutes. During interval 2 the
resource allocator allocates the resource equally among the
tasks in group A and group B, providing one ?fth of the
bandwidth to each task. The Fixed QoS structure has no way
to determine which task is temporally more important due to
conditions in the ?eld. After two minutes, the tasks in group
A are now just over half completed. During interval 3, an
additional task is received, and each task in interval 3 is
assigned one sixth of the bandwidth. At the end of three
minutes the high priority task of group A is dropped. Using
the example of FIG. 8A, this would equate with tasks
initiated for individuals in DS3 failing to complete before
the individual had left the radius.
[0043] In contrast, when time value curves are applied to
the QoS paradigm, resource allocation would proceed as is
shown by data ?ow 95; FIG. 10 is a pictorial representation
of the time-value curves per task for the data ?ow of FIG.
9. During interval 1, when burst task set A is received a
resource allocator of the present invention compares the
time-value curves of each task to identify the task with
temporal priority. As shown in FIG. 10, during interval 1 the
high level task is given control of the resource. During
interval 2, new tasks are added. However, during time
interval 2 the time value curve of the mid-range task of task
group A has temporal priority. Thus the mid-level task of
group A is completed at two minutes. At interval 3, at least
a portion of the mid-level task of group B is allocated access
to the resource. Thus it can be seen that integrating time
value curves into the policy and quality of service enforce
ment decisions increases the completion rate of high value
tasks.
[0044] Time-value curves can also used to identify tasks
that may not successfully complete in their lifetime. These
tasks are hereinafter referred to as stalled tasks. Identifying
stalled tasks allows them to be removed from resource
allocation consideration, thereby reducing performance and
providing a graceful degradation of performance in the face
of congestion.
[0045] FIG. 11A illustrates how dynamic time-value curve
analysis may be used to identify stalled tasks. In FIG. 11A,
task 126 is initiated at time. However, given the temporal
priorities of the other tasks in the work?ow, it is unlikely that
the low priority task 129 will be able to complete in its
lifetime. In the present invention, the Execution Manager
identi?es stalled ?ows and removes them from the task
queue, thereby simplifying the decision making process of
the EM and reducing congestion. FIG. 11B is a pictorial
representation of the time-value curves that are under con
sideration following the abortion of the stalled task.
[0046] FIG. 12 is a ?ow diagram illustrating several
exemplary steps that may be performed by an Execution
Manager or other resource allocator, to drop task execution
Feb. 14, 2008
using time-value curve information of the present invention.
At steps 132 and 134 a time value curve is generated for each
new task as described with regards to FIGS. 2-4. At step 136
it is determined whether the added task is going to result in
the stalling of an existing task. If so, at step 138 the stalled
task is removed from the task queue.
[0047] FIG. 13 is a graph that is provided to illustrate the
performance of a TV QoS system versus that of a ?xed QoS
as the offered load increases. As shown in FIG. 13, the
number of missed targets increases linearly with the increase
in load. In contrast, the TV QoS does not begin to drop
targets until the average tra?ic approaches the offered load
(as indicated by the dashed line X in FIG. 13).
[0048] FIG. 14 is a graph that is provided to illustrate the
graceful performance degradation that can be realiZed in a
network which implements time-value curves of the present
invention to abort stalled queues. As the siZe of the packet
payload is reduced the frequency of resource accesses is
increased. Once the offered load is less than the average
tra?ic, resource congestion quickly ensues, and data stream
are dropped. The present invention uses time-value curves to
identify the most temporally relevant tasks even during
congestion, thereby resulting in graceful network perfor
mance degradation in the face of congestion as shown in
FIG. 13.
[0049] Accordingly a method and apparatus has been
shown and described which allows Quality of Service to be
controlled at a temporal granularity. Time-value curves,
generated for each task, ensure that mission resources are
utiliZed in a manner which optimiZes mission performance.
It should be noted, however, that although the present
invention has shown and described the use of time-value
curves as applied to mission work?ow tasks, the present
invention is not limited to this application; rather, it can be
readily appreciated by one of skill in the art that time-value
curves may be used to optimiZe the delivery of any resource
to any consumer by taking into account the dynamic envi
ronment of the consumer and resource. The allocation of any
resource, including network bandwidth, storage, processing
capability, etc., may bene?t from the teachings of the present
invention.
[0050] Having described various embodiments of the
invention, it will be appreciated that many of the above
?gures are ?owchart illustrations of methods, apparatus
(systems) and computer program products according to an
embodiment of the invention. It will be understood that each
block of the ?owchart illustrations, and combinations of
blocks in the ?owchart illustrations, can be implemented by
computer program instructions. These computer program
instructions may be loaded onto a computer or other pro
grammable data processing apparatus to produce a machine,
such that the instructions which execute on the computer or
other programmable data processing apparatus create means
for implementing the functions speci?ed in the ?owchart
block or blocks. These computer program instructions may
also be stored in a computer-readable memory that can direct
a computer or other programmable data processing appara
tus to function in a particular manner, such that the instruc
tions stored in the computer-readable memory produce an
article of manufacture including instruction means which
implement the function speci?ed in the ?owchart block or
blocks. The computer program instructions may also be](https://image.slidesharecdn.com/us20080040630-141031124919-conversion-gate02/85/Time-value-curves-to-provide-dynamic-QoS-for-time-sensitive-file-transfers-19-320.jpg)
![US 2008/0040630 A1
loaded onto a computer or other programmable data pro
cessing apparatus to cause a series of operational steps to be
performed on the computer or other programmable appara
tus to produce a computer implemented process such that the
instructions Which execute on the computer or other pro
grammable apparatus provide steps for implementing the
functions speci?ed in the ?owchart block or blocks.
[0051] Those skilled in the art should readily appreciate
that programs de?ning the functions of the present invention
can be delivered to a computer in many forms; including, but
not limited to: (a) information permanently stored on non
Writable storage media (eg read only memory devices
Within a computer such as ROM or CD-ROM disks readable
by a computer l/O attachment); (b) information alterably
stored on Writable storage media (eg ?oppy disks and hard
drives); or (c) information conveyed to a computer through
communication media for example using baseband signaling
or broadband signaling techniques, including carrier Wave
signaling techniques, such as over computer or telephone
netWorks via a modem
[0052] The above description and ?gures have included
various process steps and components that are illustrative of
operations that are performed by the present invention.
These components may be implemented in hardWare, soft
Ware or a combination thereof. HoWever, although certain
components and steps have been described, it is understood
that the descriptions are representative only, other functional
delineations or additional steps and components can be
added by one of skill in the art, and thus the present
invention should not be limited to the speci?c embodiments
disclosed. In addition it is understood that the various
representational elements may be implemented in hardWare,
softWare running on a computer, or a combination thereof.
[0053] While the invention is described through the above
exemplary embodiments, it Will be understood by those of
ordinary skill in the art that modi?cation to and variation of
the illustrated embodiments may be made Without departing
from the inventive concepts herein disclosed. Accordingly,
the invention should not be vieWed as limited except by the
scope and spirit of the appended claims.
1. A method of allocating a resource to a plurality of
resource consumers includes the steps of:
generating a time-value curve de?ning a temporally
dynamic priority of a resource consumer over a time
period; and
using the time-value curve to allocate access to the
resource to the consumer over the time period.
2. The method of claim 1 Wherein the temporally dynamic
priority de?nes a temporally dynamic quality of service to
provide to the resource consumer.
3. The method of claim 1 Wherein each of the plurality of
resource consumers has an associated time-value curve and
Wherein at least tWo of the time value curves are different.
4. The method of claim 3 Wherein the step of using the
time value curve to allocate access to the resource includes
the steps of:
Feb. 14, 2008
sampling time value curves at periodic intervals to obtain
sampled time-values for each of the plurality of
resource consumers; and
using the sampled time values to select one of the plurality
of resource consumers as an access consumer for
accessing the resource for a periodic interval.
5. The method of claim 4 Wherein the time value curves
are differently Weighted.
6. The method of claim 4 Wherein the time value curves
have different lifetimes.
7. A method of allocating a resource to a plurality of
resource tasks in a Work?oW includes the steps of:
generating a time value curve for each task in the Work
How, the time-value curve de?ning a temporally
dynamic quality of service to be provided to the task,
and
using the time-value curve to select one of the plurality of
tasks.
8. The method of claim 7 Wherein the time-value curves
are associated With dynamic tasks Which have a lifetime.
9. The method of claim 8 further including the step of
using the time-value curves to identify stalled tasks.
10. The method of claim 9 including the step of aborting
stalled tasks to reduce congestion.
11. The method of claim 7 Wherein the step of generating
a time value-curve for each task includes the steps of:
selecting a generic task pro?le; and
modifying the generic task pro?le With task speci?c
information to provide a task speci?c time-value curve.
12. The method of claim 11 Wherein the generic task
pro?le is de?ned by the beloW equation I:
P(T)=Min(mT+b, MaxP)
Equation I
Wherein P is a priority of the task at time T, m is derived
through a combination of a value of the task, a lifetime
of the task and an execution time of the task, b is
determined according to the value of the task and Max
(P) represents the highest priority to allocate to the task.
13. The method of claim 7 Wherein the time-value curve
is used to calculate a quality of service for the task.
14. A policy engine comprising:
generic task pro?le information; and
time value curve generation logic operably coupled to
receive task speci?c information and to calculate a
time-value curve for the task using the generic task
pro?le information and the task speci?c information.
15. The policy engine of claim 14 Wherein the time-value
curve de?nes a temporal quality of service for the task.
16. A device comprising a computer readable medium
accessible by a resource allocator, the computer readable
medium for storing a data structure comprising, for each one
of a plurality of enquired tasks associated With the resource,
a time value curve de?ning a temporal priority to provide to
the associated task, the data structure being accessed by the
resource controller to control access to the resource.
* * * * *](https://image.slidesharecdn.com/us20080040630-141031124919-conversion-gate02/85/Time-value-curves-to-provide-dynamic-QoS-for-time-sensitive-file-transfers-20-320.jpg)
Time-value curves to provide dynamic QoS for time sensitive file transfers
- 1. US 20080040630Al (19) United States (12) Patent Application Publication (10) Pub. No.: US 2008/0040630 A1 Travostino et al. (43) Pub. Date: Feb. 14, 2008 (54) TIME-VALUE CURVES TO PROVIDE DYNAMIC QOS FOR TIME SENSITIVE FILE TRANSFERS (75) Inventors: Franco Travostino, Arlington, MA (US); Tal Lavian, Sunnyvale, CA (US); Bruce Scho?eld, Tyngsboro, MA (US); Inder Monga, Acton, MA (US) Correspondence Address: McGUINNESS & MANARAS LLP 125 NAGOG PARK ACTON, MA 01720 (US) (73) (21) (22) Assignee: Nortel Networks Limited Appl. No.: 11/469,404 Filed: Aug. 31, 2006 Related US. Application Data (60) Provisional application No. 60/721,757, ?led on Sep. Publication Classi?cation (51) Int. Cl. G06F 11/00 (2006.01) (52) US. Cl. .............................................................. .. 714/11 (57) ABSTRACT A method and apparatus has been shown and described which allows Quality of Service to be controlled at a temporal granularity. Time-value curves, generated for each task, ensure that mission resources are utiliZed in a manner which optimizes mission performance. It should be noted, however, that although the present invention has shown and described the use of time-value curves as applied to mission Work?oW tasks, the present invention is not limited to this application; rather, it can be readily appreciated by one of skill in the art that time-value curves may be used to optimize the delivery of any resource to any consumer by taking into account the dynamic environment of the con 29, 2005. sumer and resource 1 ' Command Center 2 | Y 12 Goal to Service _>D15COV‘3TY — Tranclatinn l Init Engine and Q . . Scheduler 10L Mission Resource Manag r . . Directory Execution Constraint Engine 14 Analgzcf E 1_ ‘ 1‘ t ‘ *’ Execution T ' Tender " a , ~ . Network Services Federation - lixeeution p 11 T Federation 0 O '33 _ ‘ MmmgCr (EM) Eng“? 1' Manager (NSM) 4 12 A 26 L _ v Software Exscution 0 policy Network Services Controlled v v Manzlgcr (EM) Engine 4' Manager (NSM) ' NClWUIk Network 7 1A 22 A g A A Network Resource “ _| Resoul ce lvlzmagei _ (NRM) 08 And Momtor, Software ‘ “_ Controlled " v , , , Network Network Vutualrzatroihg Resource Managw N?lwm‘k Resource L lNRM) 1 {,8 And Monitor, 35 v _ i 7 v m k’ "5' E‘xerrutznn n ‘ ‘ t 21% E
- 2. Patent Application Publication Feb. 14, 2008 Sheet 1 of 14 US 2008/0040630 A1 ~ , 7 I v v )lscovery Nelvwus Network V irrualiZation-Q Resource Manager V El ) l Command Center i Y 12 Goal to Service _> Dlscovery — Trnnulu?nn llnit Englne and Q . . Scheduler l0L MlSSlOIl Resource Manag r . . Directo Exeeutlon Constraint ry Engine 14 Anal%zcr E l_ ‘l 1‘ T ‘ i’ I _ J" v V Tender [xgmmlm ' r r Network Services Federation ~ - luxecullon P01“; 1 4 I'wlerutum 0 y 4_ M _ M M‘, (, EM ' We“ (N5 J "ringed ) 22 Engmli4 26 Z? k _ v Software Execution 4’ POncy Network Services _ Controlled " " MamagCr (EM) Engine 4- Manager (NSM) New urk 122 Q 12_6 Network 7 2 ) Remurcc Mama er Network Resource L NkM) gzs And Monitor, Software L _ Topology 7 Controlled Network Resource (NRM) E And Monitor, k’ AN Execution Federation 25 Figure 1
- 3. Patent Application Publication Feb. 14, 2008 Sheet 2 0f 14 US 2008/0040630 A1 Resource Feedback (From NRM 28, NSM 26) Work?ow Work?ow Constraints (From EM 22) (From CA 15) 24 I Policy Engine l V ‘ Work?ow a Parser 'Actor 'Task i Task constraints Generic _ Task Time-Value Curve Pro?les ‘ Generation Logic Q E V TV Augmented Work?ow ( T 0 EM 22 ) Figure 2
- 4. US 2008/0040630 A1 />30Pm5»i.9m,w>% w m Patent Application Publication Feb. 14, 2008 Sheet 3 0f 14 0E: 3095,? k 3MPm5ma9mw% w m 03? aUaGsmws.mww> m i was :aw
- 5. Patent Application Publication Feb. 14, 2008 Sheet 4 0f 14 US 2008/0040630 A1 42 Select Generic @- Task Pro?le l t Calculate time-value curve using task speci?c variables and constraints With selected generic task pro?le l 46 Store time-value curve with task Figure 4
- 6. Patent Application Publication Feb. 14, 2008 Sheet 5 0f 14 US 2008/0040630 A1 Pmi-ww m D HWl U>IvH@ m m 23:925m:50 : 0 85M0m3mMr w a. N/ w>m?so3Himsd ?
- 7. Patent Application Publication Feb. 14, 2008 Sheet 6 0f 14 US 2008/0040630 A1 435TTT T(iTme) F6igur e (<1) ‘1110M
- 8. Patent Application Publication Feb. 14, 2008 Sheet 7 0f 14 US 2008/0040630 A1
- 9. Patent Application Publication Feb. 14, 2008 Sheet 8 0f 14 US 2008/0040630 A1 5Em32M$? owqam A i
- 10. Patent Application Publication Feb. 14, 2008 Sheet 9 0f 14 US 2008/0040630 A1 80 Burst Traf?c ?liSSt’d T argcts A/3 A/3+B/3 13/3 +C/3 B/3+C/3 (/3 Using TCiumrev eVsaQ lf0uo5er Al :2X Figure 9
- 11. Patent Application Publication Feb. 14, 2008 Sheet 10 0f 14 US 2008/0040630 A1 2Interval 3Interval l In m H v-t PT: m E 8 E a F1 0 5-1 5 b1) 0 LE
- 12. Patent Application Publication Feb. 14, 2008 Sheet 11 0f 14 US 2008/0040630 A1
- 13. Patent Application Publication Feb. 14, 2008 Sheet 12 0f 14 US 2008/0040630 A1 22&5 w$Qo3mE8oakmDs m (M5 v5iwmw3roawm sm $3503 02V 02 @ OM“HmQOOHQPOHHGmSOOLUI
- 14. Patent Application Publication Feb. 14, 2008 Sheet 13 0f 14 US 2008/0040630 A1 10@ CD MTiasrsgeedt s +TV Curve Fixed QoS 7500 7250 7000 6750 6500 6250 6000 5750 5500 5250 5000 4750 4500 4250 4000 3750 3500 3250 3000 Offered Load (Kb/s) Figure 13
- 15. US 2008/0040630 A1 Patent Application Publication Feb. 14, 2008 Sheet 14 0f 14 23&5 302$ 2. :. . ~ L26= :2 __ _ wmQO=>00HE3IE .EQ5|x+ a |5m _ J2; _ _ 2>01:+i4“> cmw sumans ewq passgw
- 16. US 2008/0040630 A1 TIME-VALUE CURVES TO PROVIDE DYNAMIC QOS FOR TIME SENSITIVE FILE TRANSFERS RELATED APPLICATIONS [0001] This application claims priority under 37 CPR. §l.ll9(e) to provisional patent application 60/721,757 ?led Sep. 29, 2005 and incorporated herein by reference. FIELD OF THE INVENTION [0002] This invention relates generally to resource man agement and more particularly to a method and apparatus for generating and using time-value curves for resource man agement. BACKGROUND OF THE INVENTION [0003] The effectiveness of any mission is heavily reliant upon the ability of an underlying infrastructure to respond to the dynamic requirements of the mission. Typically missions are layered upon an existing resource infrastructure such that the mission becomes merely a set of tasks that is supported by the infrastructure. Layering a mission upon an existing infrastructure typically mis-utiliZes key resources and increases the di?iculty in detecting performance degradation or partial failures that adversely affect the mission. Allocat ing speci?c resources to a mission is technically challenging and error prone. It Would be desirable to identify a mission architecture Which Would overcome the problems of the prior art. SUMMARY OF THE INVENTION [0004] According to one aspect of the invention a method of allocating a resource to a plurality of resource consumers includes the step of generating a time-value curve de?ning a temporally dynamic priority of a resource consumer over a time period and using the time-value curve to allocate the resource to the consumer over the time period. In one embodiment, the resource is communication bandWidth, and the resource consumers include one or more tasks in a mission Work?oW. Such an arrangement permits the sched uling, pre-empting and trade-off of bandWidth betWeen different mission tasks to optimiZe deployment of mission tasks and concomitantly optimiZe mission performance. [0005] According to another aspect of the invention, a method of allocating a resource to a plurality of resource consumers includes the steps of generating a time value curve for each one of a plurality of tasks in a Work?oW, the time-value curve de?ning a temporally dynamic quality of service to be provided to the task. [0006] According to a further aspect of the invention, a policy engine comprises generic task pro?le information and time value curve generation logic operably coupled to receive task speci?c information and to calculate a time value curve for the task using the generic task pro?le information and the task speci?c information. [0007] According to a further aspect of the invention, a device comprising a computer readable medium accessible by a resource allocator is described. The computer readable medium stores a data structure comprising, for each one of a plurality of enqueued tasks associated With the resource, a time value curve de?ning a temporal priority to provide to the associated task, the data structure being accessed by the Feb. 14, 2008 resource controller to control access to the resource. These and other aspects of the invention Will be described With regard to the attached ?gures. BRIEF DESCRIPTION OF THE FIGURES [0008] FIG. 1 is a block diagram of a mission architecture in Which the present invention may be implemented; [0009] FIG. 2 is a block diagram illustrating several com ponents that may be included in a policy engine to generate time-value curves of the present invention; [0010] FIGS. 3A-3H are diagrams illustrating different types of time-value curves that may used as a basis for time-curve generation by a policy engine of the present invention; [0011] FIG. 4 is a How diagram illustrating exemplary steps that may be performed by the policy engine of FIG. 2 to generate time-value curves for one or more tasks in a Work?oW; [0012] FIG. 5 is a diagram of an exemplary netWork or service resource manager Which may use the time value curves of the present invention to control allocation of resources to Work?oW tasks; [0013] FIG. 6 is a timing diagram illustrating several exemplary time-value curves that may be assigned to tasks in a Work?oW; [0014] FIGS. 7A and 7B are timing diagrams provided to illustrate the addition of time value curves as tasks are added to a Work?oW; [0015] FIGS. 8A and 8B are used to describe an embodi ment of the present invention supporting a surveillance Work?oW; [0016] FIG. 9 is a processing timing diagram Which pic torially illustrates the bandWidth smoothing capabilities of time-value QoS servicing of a Work?oW; [0017] FIG. 10 is a pictorial representation of task selec tion using time-value curves in the example of FIG. 9; [0018] FIGS. 11A and 11B are timing diagrams used to illustrate a stalled task drop capability of the present inven tion; [0019] FIG. 12 is a How diagram illustrating exemplary steps that may be performed by a policy engine performing task dropping such as that illustrated in FIGS. 11A and 11B; [0020] FIG. 13 is a graph illustrating the performance of TV QoS versus Fixed QoS as the offered load increases; and [0021] FIG. 14 is a graph illustrating hoW TV QoS pro vides for graceful netWork failure. DETAILED DESCRIPTION [0022] FIG. 1 illustrates a mission architecture 1 in Which the present invention may advantageously be used to opti miZe mission resource utiliZation. In one embodiment the mission architecture 1 is an adaptive netWork that uses Work?oW Locked Loops (WLL) for mission resource selec tion and access management. The embodiment of FIG. 1 includes a Mission Resource Manager (MRM) 10 and a SoftWare Controlled NetWork (SCN) 12. The MRM 10 essentially orchestrates the use of mission resources by a
- 17. US 2008/0040630 A1 mission by identifying mission tasks and allocating mission resources to the mission tasks. Although the MRM may be described beloW as including certain components, it should be understood that an MRM is not a required to implement the present invention. For example, a set of Mission Goals statements may be input into the Policy Engine to help form the Time-Value curves via an administrator at a GUI or automated via a machine interface or even a Web Services Interface. [0023] An exemplary MRM 10 includes a Goal to Service translation Unit 12. The Goal to Service translation Unit may be a semantic tool Which maps a high level goal de?nition, such as a natural language de?nition, into real language mission goals into policy statements. One exem plary method for performing Goal to Service translation is described in Ser. No. (Attorney Docket No.: 120 505), entitled Mission Goal Statement to Policy Statement Translation, by Travostino et al, ?led , incorporated herein by reference. [0024] Policy statements are translated into Work?oWs using the Goal to Service Translation Unit (GSTU) 12, the Constraint Analyzer (CA) 15, the Discovery Engine and Scheduler (DES) 13 and the Directory 16. Together these components identify, quantify and qualify services that are reserved for a mission as described in patent application Ser. No. (Attorney Docket No. 120-506) for Work?oW Locked Loops to Enable Adaptive NetWorks, by Travostino et al, ?led . [0025] For example in one embodiment the directory 16 identi?es netWork resources and services that may be allo cated to missions. The DES 13 identi?es resources and services that are available for use as an infrastructure of the mission. The CA 15 evaluates the infrastructure identi?ed by the DES 13 in vieW of mission constraints. The CA 15 and DES 13 Work in concert With the goal translator to identify the mission infrastructure. An Engine uses the infrastructure and mission information to generate mission Work?oWs, Which are forWarded to the SCN 14. [0026] The SCN 14 generally controls the execution of the mission and the use of the mission resources during execu tion of the mission. The SCN 14 includes an Execution Manager (EM) 22, a Policy Engine (PE) 24, a NetWork Services Manager (26), a NetWork Resource Manager (NRM) 28 and a NetWork Resource and Topology Discovery resource 27. Together these components orchestrate the execution of the mission Work?oW using the identi?ed mission resources. [0027] For the purpose of this application a mission Work How is a collection of tasks that are performed by a collec tion of actors, Wherein an actor may be any entity that uses a mission dedicated resource. A mission dedicated resource is any resource that is at least partially dedicated to a mission. [0028] Mission tasks are dynamic, and therefore during the course of the mission different tasks are initiated and terminated. At any given time during the mission, mission tasks may seek access to a shared mission resource. Accord ing to one aspect of the invention, each task has an associ ated time-value curve indicating a temporal priority of the task over the task lifetime. The time-value curves are used during resource allocation to intelligently allocate resources Feb. 14, 2008 to tasks to optimiZe the performance of the mission. Each point in a time-value curve identi?es a priority of QoS to be associated With the task for the point in time. This alloWs QoS to be adjusted in real time based on conditions in the ?eld, thereby adding the degree of temporal granularity to the QoS structure. Controlling temporal granularity of QoS helps resolve the con?icts of congestion and While supplying all the resources to the mission Within the mission time frame, Without having to expand the siZe of the netWork. [0029] In one embodiment the PE 24 generates a time value curve for each mission task. The time value curves are forWarded to the EM 22. At resource allocation intervals the EM references the relative values of the time-value curve of each tasks that are active at that time instance to retrieve a time-value for each task, Wherein an active task is any task that has been initiated and not terminated. Using any one of a variety of selection methods, such as highest value, highest Weighted value, etc., the time values are compared to select one of the active tasks. [0030] FIG. 2 illustrates several exemplary components that may be included in a PE to generate time-value curves for tasks. The PE 24 includes a Work?oW parser 31, a generic task pro?le store 34, and time-value (TV) curve generation logic 32. The Work?oW parser receives Work?oW tasks from the Execution Machine (EM). The task may have been initiated in response to a mission trigger; for example, in a surveillance mission, a mission trigger may be the capturing of an image at a sensor point in a surveillance area. The tasks may be triggered by partial or thorough completion of another task. HoWever the task is triggered, it is forWarded to the Work?oW parser 31 Which extracts task information that may be used to build the time-value curve. Any obtain able information Which may affect a priority or quality of service to provide to the task may be provided to the time-value curve generation logic, including, for example, information regarding an actor of the task, a location of the task, a security level of the task, constraints of the tasks, dependencies of the task, poWer requirements of the task, execution time for the task, etc. Information regarding services and netWork elements that are associated With the task may also be used When calculating the time-value curve. The present invention is in no Way limited to the use of any particular task information in generating a time-value curve for priority scheduling of the task. [0031] The generic task pro?le store 34 may be used to store pre-determined task pro?les. The predetermined task pro?les are tailored to the resource requirements of various pre-de?ned tasks. For example, a surveillance task that performs high resolution image capture Would have a time value curve that temporally aligns the resource priority With the resource usage characteristics of the task. The resource requirements of a task may be selected from a set of pre-de?ned pro?les, or obtained by monitoring and pro?ling resource usage in advance to generate a time-value pro?le for the task. FIGS. 3A-3H illustrates several representative time value curve pro?les. A time value-curve pro?le such as that of FIGS. 3A and 3B may be applied to a speci?c task identi?ed as having a relatively ?at QoS. Pro?les 3C and 3D may be used to apply a linear QoS to a task. Impulse pro?les 3E and 3E may be used to provide high priority to a task for a short time. Time value curves of 7G and 7H illustrate curves that are constructed for a particular QoS pro?le.
- 18. US 2008/0040630 A1 [0032] FIG. 4 illustrates an exemplary process that may be performed by the PE 24 to construct a time-value curve for a task. At step 42 a generic pro?le associated With the task is retrieved from the pro?le store 34. At step 44, the time-value curve is piece-Wise constructed using the base pro?le in conjunction With task speci?c variables, con straints and resource information using techniques knoWn to those in the art. At step 46 the time-value curve is stored With the task. The TV augmented task is then forWarded to the EM 22. [0033] FIG. 5 illustrates exemplary components that may be included in an Execution Manager 22 that uses the time-value curves of the invention. The EM 22 Would generally include task queues 52 for storing active mission tasks. The task queues may be organiZed by mission resource. A resource controller 56 may include or otherWise have access to the Time-Value curves for the tasks, for example in a time-value queue associated With the task queue. At each mission resource allocation instance the resource controller examines the time-values of each active task associated With a resource at that instance, and selects one of the tasks based on the time-value curve. [0034] FIG. 6 is a timing diagram illustrating illustrative time value curves for 3 tasks A (62), B (64) and C (66). The generic task pro?le for each of the curves is shoWn by beloW Equation I: P(T)=Min(mT+b, MaxP) Equation I [0035] Where P is the priority, or quality of service, associated With the task at time T, m is derived based on the value of the task, the lifetime of the task and the execution time of the task and b is determined according to a value of the task. The ?at portion on each of the time-value curves in this embodiment represents the time needed to complete the task. Max (P) represents the highest priority to allocate to the task. [0036] Thus in FIG. 6 time value curve C is a loWer value task, Which takes a relatively short period of time to com plete. Curve B has the highest initial priority task, With a shorter lifetime than curve C. Curve A has the second highest initial priority, but the shortest lifetime. Using strict priority allocation technique, task B Will obtain access to the resource until time T1. At time T1, access begins to be granted to task A, With task A reaching its maximum priority at time 2 to alloW it to be at maximum priority for the time period associated With executing the task. The time-value curves may thus be used provide a temporal priority of the task Which ensures that the task can complete in its lifetime. [0037] FIGS. 7A and 7B are pictorial representations of a mission Work?oW, provided to illustrate the dynamic nature of Work?oW tasks; tasks are continually instantiated and terminated during the Work?oW. For example, at time T4, a loW priority task 68 is triggered. At time T7 a higher priority task is triggered. At any resource allocation period, the time value curves may be used to real-time, dynamic resource management for a dynamic mission Work?oW. [0038] FIGS. 8A and 8B Will be used to describe an exemplary embodiment of the present invention for use in a particular mission. The mission of FIG. 8 is a surveillance mission Which seeks to identify individuals Within a de?ned radius. The mission comprises dependent tasks, With a ?rst set of tasks being associated With the detection and collec Feb. 14, 2008 tion of information for each an individual that comes Within a radius, a second set of tasks being associated With col lecting high resolution information regarding individuals Within the radius, a third set of tasks being associated With collecting increasingly images at very high resolution. Each of the tasks of in the mission use netWork bandWidth for the exchange of information associated With the target. The individuals are grouped into loW value targets (initial prior ity 0); mid value targets (initial priority value 1) and high value targets (initial priority value 2). The targets are detected using mission resource sensors distributed Within the radius. A detection of the target results in an attempt to collect information about the target to determine if the target is a target of interest. The collection of information may include the transmission of identi?cation information, such as a loW resolution or high resolution image. The type of collection to be performed for each target is a task in the above described architecture. The task is assigned a time value curve Which provides a temporal priority to the collection of information for the particular target. [0039] The time value curves of FIG. 8B are associated With tasks to be performed on three targets. A ?rst task 72 is associated With a mid priority target Which has been sensed in area DS3. There are only 3 minutes to collect information regarding the target before the target leaves the area. Thus the time-value curve is adapted for this target to ensure that data collection for this target can occur Within the target tracking task lifetime. A second task 74 is associated With a target in area DSa. The target has been identi?ed as a high value target, perhaps as a result of a loW-resolution image mapping task. The target is exiting the area in 4 minutes. The slope and intercept of the time-value curve re?ect the perceived target value and detected target location, and ensures that adequate processing of the target may be completed before the target is lost. Time value curve 76 is associated With a task generated in response to target detec tion in area DS2b. There are 5 minutes until the target quits the area, and thus time-value curve 76 re?ects the loWer quality of service level to be provided to the collection of information regarding this target. [0040] FIG. 9 is a data How diagram intuitively illustrating the bene?ts of using the time-value curves of the present invention to provide real-time temporally prioritized resource allocation in to mission tasks. As Will be seen from the beloW description, the time-value curves e?fectively smooth and optimiZe resource allocation among the tasks. Optimizing the resource allocation improves the completion rate of mission tasks and increases the chances of mission success. [0041] Data series 80 illustrates a stream of resource requests that are received in a data stream. The performance of the mission depends upon being able to process each task in the data stream during its lifetime; if tasks are dropped or delayed, mission information is lost. In FIG. 9, each data slot represents a one minute interval, and assume that the Burst tra?ic received in interval 1 includes high, medium and loW priority tasks having respective lifetimes of three, four and ?ve minutes. [0042] Data How 90 illustrates an exemplary result of resource allocation for a mission using Fixed QoS resource allocation methods. In a typical Fixed QoS resource alloca tion method, a ?xed bandWidth percentage is allocated to
- 19. US 2008/0040630 A1 each task. During interval 1 of Data ?ow 90, Burst Tra?ic for three tasks is received. The prior art resource allocator, receiving no information from the ?eld regarding a temporal priority of the tasks, uses a Fixed QoS to allocate resources to the tasks. Each of the three tasks is provided one third of the bandwidth. At the end of interval 1 (or one minute into the task), each of the three pending tasks in group A is 1/3 complete. During interval 2, an additional task group B is received which includes two additional tasks. Assume group B includes a medium and high priority task with respective lifetimes of four and ?ve minutes. During interval 2 the resource allocator allocates the resource equally among the tasks in group A and group B, providing one ?fth of the bandwidth to each task. The Fixed QoS structure has no way to determine which task is temporally more important due to conditions in the ?eld. After two minutes, the tasks in group A are now just over half completed. During interval 3, an additional task is received, and each task in interval 3 is assigned one sixth of the bandwidth. At the end of three minutes the high priority task of group A is dropped. Using the example of FIG. 8A, this would equate with tasks initiated for individuals in DS3 failing to complete before the individual had left the radius. [0043] In contrast, when time value curves are applied to the QoS paradigm, resource allocation would proceed as is shown by data ?ow 95; FIG. 10 is a pictorial representation of the time-value curves per task for the data ?ow of FIG. 9. During interval 1, when burst task set A is received a resource allocator of the present invention compares the time-value curves of each task to identify the task with temporal priority. As shown in FIG. 10, during interval 1 the high level task is given control of the resource. During interval 2, new tasks are added. However, during time interval 2 the time value curve of the mid-range task of task group A has temporal priority. Thus the mid-level task of group A is completed at two minutes. At interval 3, at least a portion of the mid-level task of group B is allocated access to the resource. Thus it can be seen that integrating time value curves into the policy and quality of service enforce ment decisions increases the completion rate of high value tasks. [0044] Time-value curves can also used to identify tasks that may not successfully complete in their lifetime. These tasks are hereinafter referred to as stalled tasks. Identifying stalled tasks allows them to be removed from resource allocation consideration, thereby reducing performance and providing a graceful degradation of performance in the face of congestion. [0045] FIG. 11A illustrates how dynamic time-value curve analysis may be used to identify stalled tasks. In FIG. 11A, task 126 is initiated at time. However, given the temporal priorities of the other tasks in the work?ow, it is unlikely that the low priority task 129 will be able to complete in its lifetime. In the present invention, the Execution Manager identi?es stalled ?ows and removes them from the task queue, thereby simplifying the decision making process of the EM and reducing congestion. FIG. 11B is a pictorial representation of the time-value curves that are under con sideration following the abortion of the stalled task. [0046] FIG. 12 is a ?ow diagram illustrating several exemplary steps that may be performed by an Execution Manager or other resource allocator, to drop task execution Feb. 14, 2008 using time-value curve information of the present invention. At steps 132 and 134 a time value curve is generated for each new task as described with regards to FIGS. 2-4. At step 136 it is determined whether the added task is going to result in the stalling of an existing task. If so, at step 138 the stalled task is removed from the task queue. [0047] FIG. 13 is a graph that is provided to illustrate the performance of a TV QoS system versus that of a ?xed QoS as the offered load increases. As shown in FIG. 13, the number of missed targets increases linearly with the increase in load. In contrast, the TV QoS does not begin to drop targets until the average tra?ic approaches the offered load (as indicated by the dashed line X in FIG. 13). [0048] FIG. 14 is a graph that is provided to illustrate the graceful performance degradation that can be realiZed in a network which implements time-value curves of the present invention to abort stalled queues. As the siZe of the packet payload is reduced the frequency of resource accesses is increased. Once the offered load is less than the average tra?ic, resource congestion quickly ensues, and data stream are dropped. The present invention uses time-value curves to identify the most temporally relevant tasks even during congestion, thereby resulting in graceful network perfor mance degradation in the face of congestion as shown in FIG. 13. [0049] Accordingly a method and apparatus has been shown and described which allows Quality of Service to be controlled at a temporal granularity. Time-value curves, generated for each task, ensure that mission resources are utiliZed in a manner which optimiZes mission performance. It should be noted, however, that although the present invention has shown and described the use of time-value curves as applied to mission work?ow tasks, the present invention is not limited to this application; rather, it can be readily appreciated by one of skill in the art that time-value curves may be used to optimiZe the delivery of any resource to any consumer by taking into account the dynamic envi ronment of the consumer and resource. The allocation of any resource, including network bandwidth, storage, processing capability, etc., may bene?t from the teachings of the present invention. [0050] Having described various embodiments of the invention, it will be appreciated that many of the above ?gures are ?owchart illustrations of methods, apparatus (systems) and computer program products according to an embodiment of the invention. It will be understood that each block of the ?owchart illustrations, and combinations of blocks in the ?owchart illustrations, can be implemented by computer program instructions. These computer program instructions may be loaded onto a computer or other pro grammable data processing apparatus to produce a machine, such that the instructions which execute on the computer or other programmable data processing apparatus create means for implementing the functions speci?ed in the ?owchart block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing appara tus to function in a particular manner, such that the instruc tions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function speci?ed in the ?owchart block or blocks. The computer program instructions may also be
- 20. US 2008/0040630 A1 loaded onto a computer or other programmable data pro cessing apparatus to cause a series of operational steps to be performed on the computer or other programmable appara tus to produce a computer implemented process such that the instructions Which execute on the computer or other pro grammable apparatus provide steps for implementing the functions speci?ed in the ?owchart block or blocks. [0051] Those skilled in the art should readily appreciate that programs de?ning the functions of the present invention can be delivered to a computer in many forms; including, but not limited to: (a) information permanently stored on non Writable storage media (eg read only memory devices Within a computer such as ROM or CD-ROM disks readable by a computer l/O attachment); (b) information alterably stored on Writable storage media (eg ?oppy disks and hard drives); or (c) information conveyed to a computer through communication media for example using baseband signaling or broadband signaling techniques, including carrier Wave signaling techniques, such as over computer or telephone netWorks via a modem [0052] The above description and ?gures have included various process steps and components that are illustrative of operations that are performed by the present invention. These components may be implemented in hardWare, soft Ware or a combination thereof. HoWever, although certain components and steps have been described, it is understood that the descriptions are representative only, other functional delineations or additional steps and components can be added by one of skill in the art, and thus the present invention should not be limited to the speci?c embodiments disclosed. In addition it is understood that the various representational elements may be implemented in hardWare, softWare running on a computer, or a combination thereof. [0053] While the invention is described through the above exemplary embodiments, it Will be understood by those of ordinary skill in the art that modi?cation to and variation of the illustrated embodiments may be made Without departing from the inventive concepts herein disclosed. Accordingly, the invention should not be vieWed as limited except by the scope and spirit of the appended claims. 1. A method of allocating a resource to a plurality of resource consumers includes the steps of: generating a time-value curve de?ning a temporally dynamic priority of a resource consumer over a time period; and using the time-value curve to allocate access to the resource to the consumer over the time period. 2. The method of claim 1 Wherein the temporally dynamic priority de?nes a temporally dynamic quality of service to provide to the resource consumer. 3. The method of claim 1 Wherein each of the plurality of resource consumers has an associated time-value curve and Wherein at least tWo of the time value curves are different. 4. The method of claim 3 Wherein the step of using the time value curve to allocate access to the resource includes the steps of: Feb. 14, 2008 sampling time value curves at periodic intervals to obtain sampled time-values for each of the plurality of resource consumers; and using the sampled time values to select one of the plurality of resource consumers as an access consumer for accessing the resource for a periodic interval. 5. The method of claim 4 Wherein the time value curves are differently Weighted. 6. The method of claim 4 Wherein the time value curves have different lifetimes. 7. A method of allocating a resource to a plurality of resource tasks in a Work?oW includes the steps of: generating a time value curve for each task in the Work How, the time-value curve de?ning a temporally dynamic quality of service to be provided to the task, and using the time-value curve to select one of the plurality of tasks. 8. The method of claim 7 Wherein the time-value curves are associated With dynamic tasks Which have a lifetime. 9. The method of claim 8 further including the step of using the time-value curves to identify stalled tasks. 10. The method of claim 9 including the step of aborting stalled tasks to reduce congestion. 11. The method of claim 7 Wherein the step of generating a time value-curve for each task includes the steps of: selecting a generic task pro?le; and modifying the generic task pro?le With task speci?c information to provide a task speci?c time-value curve. 12. The method of claim 11 Wherein the generic task pro?le is de?ned by the beloW equation I: P(T)=Min(mT+b, MaxP) Equation I Wherein P is a priority of the task at time T, m is derived through a combination of a value of the task, a lifetime of the task and an execution time of the task, b is determined according to the value of the task and Max (P) represents the highest priority to allocate to the task. 13. The method of claim 7 Wherein the time-value curve is used to calculate a quality of service for the task. 14. A policy engine comprising: generic task pro?le information; and time value curve generation logic operably coupled to receive task speci?c information and to calculate a time-value curve for the task using the generic task pro?le information and the task speci?c information. 15. The policy engine of claim 14 Wherein the time-value curve de?nes a temporal quality of service for the task. 16. A device comprising a computer readable medium accessible by a resource allocator, the computer readable medium for storing a data structure comprising, for each one of a plurality of enquired tasks associated With the resource, a time value curve de?ning a temporal priority to provide to the associated task, the data structure being accessed by the resource controller to control access to the resource. * * * * *