Adaptive congestion control protocol (accp) for wireless sensor networks

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 5, No. 5, October 2013

ADAPTIVE CONGESTION CONTROL PROTOCOL
(ACCP) FOR WIRELESS SENSOR NETWORKS
James DzisiGadze1,DelaliKwasiDake2, and Kwasi Diawuo3
1

Department of Electrical & Electronic Engineering, Kwame Nkrumah University of
Science and Technology, Kumasi, Ghana
2
Department of Computer Engineering, Kwame Nkrumah University of Science and
Technology, Kumasi, Ghana
3
Department of Computer Engineering, Kwame Nkrumah University of Science and
Technology, Kumasi, Ghana

ABSTRACT
In Wireless Sensor Networks (WSN) when an event is detected there is an increase in data traffic that might
lead to packets being transmitted through the network close to the packet handling capacity of the WSN.
The WSN experiences a decrease in network performance due to packet loss, long delays, and reduction in
throughput. In this paper we developed an adaptive congestion control algorithm that monitors network
utilization and adjust traffic levels and/or increases network resources to improve throughput and conserve
energy. The traffic congestion control protocol DelStatic is developed by introducing backpressure
mechanism into NOAH. We analyzed various routing protocols and established that DSR has a higher
resource congestion control capability. The proposed protocol, ACCP uses a sink switching algorithm to
trigger DelStatic or DSR feedback to a congested node based on its Node Rank. From the simulation
results, ACCP protocol does not only improve throughput but also conserves energy which is critical to
sensor application survivability on the field. Our Adaptive Congestion control achieved reliability, high
throughput and energy efficiency.

KEYWORDS
Energy Efficient Congestion Control, Event Detection, Traffic Control, Resource Control, Wireless Sensor
Network.

1. INTRODUCTION
The emerging field of wireless sensor network (WSN) has potential benefits for real-time system
monitoring. A wireless sensor network consists of remotely deployed wireless sensor nodes in a
physical phenomenon. The embedded sensors continuously monitor the physical process and
transmit information in a multi-hop fashion to a special node called the sink. WSN therefore has
three basic characteristics: centralized data collection, multi-hop data transmission, and many-toone traffic patterns. It means that the nodes closer to the base stations need to send more data
packets and their traffic burden will be more severe. This leads to severe packet collisions,
network congestion, and packet loss. In most severe cases it even results in congestion
collapse.Within the framework of (WSNs), there are many application areas where sensor
networks are deployed: for environmental monitoring, battlefield surveillance, health and
industrial monitoring control. We use the recent oil find in Ghana and its associated
environmental impact as example of one application and describe congestion problem in WSNs in
this context. Residents of oil and gas field communities often report incidents of: asthma,
respiratory and cardiovascular illnesses, autoimmune diseases, liver failure, cancer and other
DOI : 10.5121/ijwmn.2013.5509

129


ailments such as headaches, nausea, and sleeplessness. These health effects could be the result of
air contaminant such as Volatile Organic Compounds (VOCs). We envisage the use of wireless
sensor network in Ghana to monitor the quality of air in oil and gas field communities. One of the
main problems to the success of the monitoring scheme might be congestion in the wireless
sensor network.
A typical WSN deployment scenario is illustrated in figure 1. The VOC measurement in the oil
and gas field is done by sensors deployed in the field. Information gathered from the sensor nodes
is transmitted in a multi-hop fashion to the sink node and then to the gateway computer as shown
in Figure1.

Figure 1: WSN deployment scenario

The traffic in the network is low under light load conditions. When an event is detected, the load
increases and transmission of traffic through network approaches the packet handling capacity of
the network. Congestion is a state in a network when the total sum of demands on network
resources is more than its available capacity. Mathematically:
∑ Demand > Available Resources (1)

In WSNs, congestion happens due to contention caused by concurrent transmission, buffer
overflows and dynamic time varying wireless channel condition [1][2][3].
At low levels of VOC pollution when the load is light, throughput and hence the wireless sensor
network utilization increases as the load increases thus before point A in figure 2a.
As the load increases, a point A in Figure 2a is reached beyond which the wireless network
utilization (throughput) increases at a rate lower than the rate the load is increased and the
wireless sensor network enters into moderate congestion state. As the load continues to increase
and the queue lengths of the various nodes continue to grow, a point B is reached beyond which
the throughput drops with increased load.

130


As seen in figure 2a, the sharp decline of throughput from point B is a state where wireless sensor
network is rendered inefficient since critical network resource such as energy and bandwidth are
wasted. This is a state where bandwidth and buffer overflow have reached a threshold value that
any further packet arrival leads to severe network performance degradation. In terms of VOCs
monitoring from the oil field, the gateway computer will receive and report data that is misleading
due to the severe congestion. It is therefore imperative to control congestion in the WSN.
Existing congestion control protocols solve the severe congestion problem using unilateral
congestion control approach [4][5][6]. That is the use of either traffic control or resource control
strategy. There exists, for instance, an inverse relationship between energy and throughput in
these approaches.
‫ܧ‬ൌ

1
, ‫ ܧ‬ൌ ݁݊݁‫ ܲܶ ,ݕ݃ݎ‬ൌ ݄ܶ‫ݐݑ݌݄݃ݑ݋ݎ‬
ܶܲ

ሺ2ሻ

A good energy and throughput performance cannot be achieved through this unilateral approach.
In this paper, we proposed the development of ACCP a protocol that uses a switch algorithm to
trigger DelStatic (traffic) or DSR (resource) protocol that results in both good energy and
throughput performance in WSNs.
The remainder of the paper is organized as follows: Section II provides review on literature,
Section III on ACCP Design, Section IV on ACCP NR and Feedback Calculation and the final
section on Simulation and Results.

2. LITERATURE REVIEW
A number of previous works have addressed the issue of congestion control in Wireless Sensor
Networks [7]. In these works congestion alleviation is either by traffic control or resource control
approach.

2.1. Traffic Control
Congestion at a node happens when the incoming traffic volume exceeds the amount of resources
available to the node. To alleviate congestion, the incoming data is throttled (referred to as traffic
control) mostly using hop-by-hop backpressure. Congestion Detection and Avoidance (CODA)
131


uses Buffer size and Channel condition to detect congestion and then reduces the rate of incoming
traffic into the network[8]. The algorithm used to adjust traffic rate works in a way like additive
increase multiplicative decrease (AIMD).
The reduction in traffic affects the accuracy level when congestion is transient. In such transient
congestion condition, increasing resources such as hop-by-hop bandwidth or additional data path
may be a better option to cope with congestion. In addition CODA used Closed Loop end-to-end
upstream approach to control congestion. This approach generally has a longer response time
when congestion occurs, and in-turn results in dropping of lots of segments [9]. The drop of
segments results in the waste of limited energy resources.
Event-to-Sink Reliable Transport (ESRT) is traffic congestion scheme that provides reliability
from sensors to sink in addition to congestion control [10]. It jointly uses average local packet
service time and average local packet inter-arrival time in order to estimate current local
congestion degree in each intermediate node. The use of hop-by-hop feedback control removes
congestion quickly.
Priority based congestion control protocol (PCCP) [11] defines a new variable, congestion degree
as a ratio of average packet service time over average packet interval arrival time at each sensor
node. PCCP uses a hop-by-hop rate adjustment technique called priority-based rate adjustment
(PRA) to adjust scheduling rate and the source rate of each sensor node in a single-path routing
WSN. The idea is to allow data flow generated by a source node to pass through the nodes and
links along with single routing path. With PCCP control, it is easy for sensor nodes to learn the
number of upstream data sources in the sub tree roots and measure the maximum downstream
forwarding rate. This helps the congested nodes to calculate the per-source rate based on priority
index of each source node.
Dynamic Predictive Congestion Control (DPCC) [12] predicts congestion in a node and
broadcasts traffic on the entire network fairly and dynamically. To achieve a high throughput
value, DPCC uses three rate adjustment techniques: backward and forward node selection (BFS),
predict congestion detection (PCD) and dynamic priority-based rate adjustment (DPRA), which
are introduced with responsibility for precise congestion discovery and weighted fair congestion
control.
In a hop by hop mitigation congestion control protocol [13], congestion detection is based on the
following parameters: buffer size, hop count and MAC overhead. When congestion is detected at
a node, each downstream node is set with a congestion bit using Node Rank threshold policy.
With hop by hop traffic control mitigation strategy, each source node will adjust their
transmission rate dynamically based on the RAF feedback from a congested node.

2.2. Resource Control
In resource congestion control schemes when data traffic at a node increases beyond the resources
available at that node, the resources to the node are increased to enable the node cope with the
excess traffic. In [14] a resource congestion control scheme is developed where large numbers of
sensor nodes are turned off during normal traffic. When congestion is detected some nodes are
woken up to form one or more additional routing paths called multiplexing paths. The congestion
is taken care of by distributing the incoming traffic over the original path and the multiplexing
paths. The challenge is that precise network resource adjustment is needed to avoid over- or
under- provision of resources. The paper established that when congestion is transient, increasing
resources by creating multiple paths around the hotspot effectively increases the number of
delivered packets (accuracy level), and saves a lot of energy by avoiding collisions and
retransmissions.
132


3. ACCP DESIGN
Sensitive wireless sensor applications such as medical and hazardous environmental monitoring
are loss-intolerant. In such applications it is required to have accurate data (high fidelity) from the
source nodes to the sink node for reliable analysis. The use of traffic congestion control in
particular during transient congestion may result in the reduction in the accuracy level of data
reaching the sink. ACCP uses a sink switching algorithm to switch between traffic congestion
control logic (DelStatic) and resource congestion control logic (DSR). DelStatic is developed by
introducing backpressure mechanism into NO Ad-Hoc Routing Agent (NOAH). We analyzed
various routing protocols and established that DSR has a higher resource congestion control
capability.
As shown in Figure 3, an intermediate node receives data traffic from multiple source nodes. The
many-to-one modelposes threat to the resources available to the intermediate nodes. If bandwidth
and buffer occupancy usage of the node exceeds a threshold, congestion sets in. It is important to
control congestion at the intermediate node to guarantee reliable transmission of data generated
by the source nodes to the sink node (base station)

Figure 3: Wireless Sensor Network and data transmission

ACCP is thus based on two mechanisms:
1) Detect congestion at a node using Buffer Occupancy & Channel Utilization Strategy
2) Control the detected congestion by initializing ACCP protocol

3.1. Congestion Detection
To effectively detect congestion, we implement in ACCP a double congestion detection
mechanism: channel utilization strategy and buffer occupancy.

3.3.1. Buffer Occupancy
In ACCP the instantaneous buffer occupancy of intermediate nodes is compared to a threshold
value. If the threshold value is reached, then congestion may be about to set in. This kind of
detection is also seen in [15][16]. In order to avoid late buffer threshold detection, the buffer
growth rate is also monitored. An exponentially weighted moving average of the instantaneous
queue length is used as a measure of congestion [17]:
‫ܩܸܣ‬௤ ൌ ൫1 െ ‫ݓ‬௤ ൯ ‫ܩܸܣ כ‬௤ ൅ ‫ݓ‬௤ ‫ݐݏ݊݅ כ‬௤

ሺ3ሻ

Where: 1ሻ ‫ܩܸܣ‬௤ ൌ ܽ‫݁݀݋݂݊݋݄ݐ݈݃݊݁݁ݑ݁ݑݍ݁݃ܽݎ݁ݒ‬
133

2ሻ ‫ݓ‬௤ = weighted queue length of node
3ሻ ݅݊ ‫ݐݏ‬௤ ൌ ‫ݐ݊ܽݐݏ݊݅ݐ݊݁ݎݎݑ݄ܿ݁ݐݐܽ݁݀݋݂݊݋ݐ݄݈݃݊݁݁ݑ݁ݑݍ‬

The average queue length is updated whenever a packet is inserted into the queue. Thus, if ‫ܩܸܣ‬௤
exceeds a certain upper threshold U, the node is said to be congested. The node remains in a
congestion state until ‫ܩܸܣ‬௤ falls below a lower threshold L. In practice, a single threshold is too
coarse-grained to effectively react to congestion. Buffer occupancy alone is not a reliable
congestion indicator because packets can be lost in the channel due to collision or hidden terminal
situations and have no chance to reach a buffer [18]. ACCP implements both Channel Utilization
Strategy and buffer occupancy for effective congestion detection.

3.3.2. Channel Utilisation and Detection Strategy
Channel Utilization is the fraction of time the channel is busy due to transmission of frames. High
channel utilization is used as an indication of congestion. When a sensor node has a packet to be
sent, it samples the state of the channel at regular interval. Based on the number of times the
channel is found to be busy, the node calculates a utilization factor which when above a certain
level indicates congestion.
Channel Utilization technique uses Carrier Sensing Multiple Access (CSMA) algorithm to listen
to the communication medium before data is transmitted. This is accomplished with the help of
the MAC Protocol. In ACCP we used IEEE 802.11 MAC with collision avoidance. The channel
occupation due to MAC contention is computed as:
Cocc=t RTS+tCTS+3t SIFS

(4)

where
tRTS and tCTS are the time spent on RTS and CTS, exchanges and t SIFS is the SIFS period.
Then the MAC overhead is represented as
OHMAC =Cocc+t acc
wheret acc is the time taken due to access contention. That is,
OHMAC is strongly related to the congestion around a given node.

(5)

The channel busy ratio is another factor used in ACCP
‫ ܴܤܪܥ‬ൌ

‫ݐ݈݊ܤ‬
ܶ‫ݐ݋ݐ‬

ሺ6ሻ

Where ‫ ݐ݈݊ܤ‬represents the time interval that the channel is busy due to successful transmission or
collision, and ܶ‫ ݐ݋ݐ‬represents the total time.
To maximize channel utilization detection strategy, the channel detection delay of wireless
medium is of essence [19].
ܵ௠௔௫ ൎ

1

൫1 ൅ 2ඥߚ൯

൬݂‫ ߚݎ݋‬ൌ

߬‫׋‬
‫1 ا‬൰
‫ܮ‬

134


1ሻ ܵ௠௔௫ = CSMA maximum theoretical throughput
approximation
2) β = the measure of radio propagation delay and channel
detection delay
3) τ = the sum of both radio propagation delay and channel
idle detection delay in seconds
4)C = the raw channel bit rate
5)L = the expected number of bits in a data packet
In achieving a high ܵ௠௔௫ value for good performance, we lowered the β value, which determines
how quickly a node can detect idle periods. CSMA efficiency is highly dependent on the β value.

3.2. ACCP Algorithm
To effectively detect congestion, we implement in ACCP a double congestion detection
mechanism: channel utilization strategy and buffer occupancy.
The NS2 modular implementation of ACCP is based on DSR and DelStatic protocols. The
Dynamic Source Routing (DSR) is an on-demand routing protocol that is based on the concept of
source routing where the sender of a packet determines the complete sequence of nodes through
which, the packets are forwarded.
Our DSR logic has the capability to reduce the sleeping interval of backup nodes near the
congested node so that they become active. One’s they become active, they can communicate
with each other using DSR Route Discovery and Route Maintenance algorithm to trace the sink
node. The goal of DRS is:
•
•

To increase resource provisioning as soon congestion occurs;
To reduce the resource budget as soon as congestion subsides

The DelStatic logic has a backpressure mechanism that allowsa sensor node that has its channel
utilization level and buffer occupancy above a threshold broadcasts a suppression message to
upstream nodes. It uses an open-loop hop-by-hop data flow strategy until the message reaches the
source nodes. Each node reacts to the suppression message by throttling its sending rate or drop
packets using packet drop and Additive Increase Multiple Decrease (AIMD) policy. The
DelStatic traffic reduction scenario is as shown in Figure 4.

135


Figure 4: DelStatic Traffic Reduction Scenario

The goal of this work is to allow DelStatic and DSR congestion control protocols to work in one
framework by implementing a switching mechanism that is controlled by the sink node. From the
framework diagram in Figure 5, congestion detection is done at the node level with a feed sent to
the sink to report congestion in the network.

Figure 5: Framework for Adaptive Protocol

The sink uses two important parameters, Energy Remaining (ER) and Sensor Application Priority
(SP) to determine the control to the sending node. Either Traffic Control (DelStatic) or Resource
Control (DSR) is triggered by the sink depending on the algorithm below:

136


Switch (congestion control){
Case 0:
If (Energy > threshold){resource control}
Case1:
If (Energy > threshold){traffic control}
Case2:
If (High- fidelity/priority application){resource control}
Case3:
If (Low- fidelity/priority application){traffic control}
default: congestion control not triggered}

3.3. ACCP NR and Feedback Calculation
With adaptive control, each node calculates its node rank (NR) based on its Buffer level and
Channel Utilization Ratio. Here the channel busy ratio represents the interference level and is
defined as the ratio of time intervals when the channel is busy due to successful transmission or
collision to the total time. After estimating its rank, a node forwards this to its downstream node.
When NR crosses a threshold T, then the node will set the congestion bit (CB) in every packet it
forwards. On receiving this node rank, the downstream node will first check for the congestion
bit. If it is not set, it will simply compute its node rank and adds it to the rank obtained from its
previous node and passes on to the next node. On the other hand, if the congestion bit is set, the
node sends a feed to the sink node to confirm whether to trigger DelStatic or DSR control.
Depending on the energy remaining to the network and the sensor application priority, the sink
will send a feedback control to the sending node. This node will in-turn based on the information
received from the sink set Rate Adjustment Feedback (RAF) either triggers an alternate path
(DSR Control) or transmit it towards the source as a feedback. On receiving the feedback, the
source nodes will adjust their transmission rate (DelStatic).
Each node calculates its node rank NR based on the following parameters BSize, HC, CBHR and
OHMAC, where BSizeis the buffer size of the node and HC is the hop count value.
ܴܰ ൌ ߙ1. ‫ ݁ݖ݅ܵܤ‬௡ଵ ൅ ߙ2. ‫ ܪ‬൅ ߙ3. ‫ ܴܤܪܥ‬൅ ߙ4 ܱ‫ܪ‬ெ஺஼

ሺ8ሻ

Here ߙ1, ߙ2, ߙ3 ܽ݊݀ ߙ4 are constant weight factors whose values between 0 and 1
Each data packet has a congestion bit (CB) in its header. Every sensor node maintains a threshold
T. When NR crosses the value of T the node will set its CB in every packet it forwards.
In Figure 6, for example nodes n1, n2, n3 estimates their rank NR1, NR2, NR3 and forwards this
to its downstream node n4 along with data packets. On receiving NR1, NR2 and NR3, n4 will
first check their CB value. If it is not set, it will simply compute its node rank NR4 and adds it to
the rank obtained from nodes n1, n2, and n3 and passes it to the next node or sink. On the other
hand, if the CB is set at n1, n2 and n3, the node n4 will send a feed to the sink about the
congestionscenario. The sink node uses energy remaining (ER) to the network and priority of
sensor application (SP) to determine which control to use for feedback

137


Figure 6: Node Rank (NR) Propagation

The sending node receives either N1 (DSR Control) or N0 (DelStatic Control) from the sink as a
feedback signal. The sink operates the feedback with request from directed diffusion protocol to
disseminate N1 and N0 information to the congested nodes.

Figure 7: Feedback Propagation

The sending node based on the N1 or N0 feedback information from sink calculates its Rate
Adjustment Feedback (RAF) based on the rank as:
ܴ‫ ܨܣ‬ൌ ൬

‫݁ݐܽݎܣ‬
൰ െ ∑ ܱ‫ܪ‬ൗ‫ ݅ܥܣܯ‬െ ∑‫ܴ݅ܪܤܥ‬
‫ܥܪ‬

ሺ9ሻ

Where ‫ ݁ݐܽݎܣ‬is the arrival rate of packets at node n which is given as:
‫ ݁ݐܽݎܣ‬ൌ

ே௉
௧

ሺ10ሻ

Here NP – is the number of packets received and t is the time of interval for the packet
transmission.
When feedback (N1) is received, the RAF is propagated to a neighboring node by waking up a
sleeping node (DSR Control) as shown in figure 7. On receiving the feedback packet, a
neighboring node n6 is signaled to continue data transmission.
138


When feedback (N0) is received, backpressure algorithm is initiated to the upstream nodes n1, n2,
n3 to adjust their transmission rate by dropping packet or triggering a delay in milliseconds. To
adjust the rate dynamically, DelStatic protocol uses formulae on node
ܰ‫ ݁ݐܽݎ‬ൌ ܰ‫ ݁ݐܽݎ‬െ ܴ‫ܨܣ‬

ሺ11ሻ

Thus the traffic rate is adaptively adjusted according to the MAC contention and buffer size. The
procedures are repeated for all the hops towards the sinks which are congested.

4. SIMULATION RESULTS
We use ns2 to simulate ACCP protocol. For the purpose of achieving our objective, we modified
the routing agent class of NOAH to ‘start-DelStatic’. The ACCP agent interface linkage to MAC
component is also modified to enable fix routing in DelStatic Control.
In testing ACCP protocol, sensor nodes of sizes 25, 64 and 150 nodes were deployed using grid
and random topology for 10 seconds of simulation time. We set all nodes to have transmission
range of 250 meters, interference range of 550 meters and maximum packet in queue of 50. The
simulated traffic source is Constant Bit Rate. The performance of DelStatic compared with
NOAH protocol using the metrics of throughput and energy consumption.
The result of the simulations is an output trace file that can be used to do data processing and to
visualize the simulation with a program called Network Animator (NAM). For this project, NS
version 2.34 is used.

4.1. Simulation Parameters
Table 1: Simulation Parameters

4.2. Energy Parameters
Set opt(initialenergy) 900
;# Initial energy in Joules
-rxPower 0.3
;# power consumption in state (watt)
-txPower 0.6
;# power consumption in sleep state (watt)
139


4.3. Part 1: TADD Protocols
Simulation is conducted on Ad hoc On-demand Vector Protocol (AODV), Destination-Sequenced
Distance Vector (DSDV), Dynamic Source Routing (DSR) and Temporarily Ordered Routing
Algorithm (TORA) which forms the TADD protocols. The aim is to compare these routing
protocols and their performance in terms of traffic bytes generated, end to end delay and
throughput. The protocol with a higher metric value, DSR is used for the resource controlling part
of ACCP.

Figure 8: Grid Topology 25 nodes [ 2 Source / 1 Sink]

In Figure 9, the average end to end delay has a maximum value at ‘5.5’ and a minimum at ‘0.0’.
This is a positive result from DSR protocol. Though 5.5 seconds is on the high side, it is an
indication that CBR source data is transmitted over a significant period. Congestion is effectively
dealt with during the transmission process.

Figure 9: End to End Delay of DSR

4.4. Part 2: NoAH and DelStatic
Our proposed traffic control ( DelStatic) protocol is compared with NOAH protocol using metrics
of end to end delay and throughput. The generated graph shows a significant throughput increase
in DelStatic protocol over a NOAH protocol.

140


It is observed in Figure 10 that the delay performance of DelStatic is better than that of NOAH.
The better delay performance is due to the backpressure algorithm we introduced in DelStatic.
The backpressure algorithm is significant in controlling congestion during data packet
transmission from source nodes to the sink node.

Figure 10: End to End Delay of DelStatic and NOAH

Throughput is the average rate of successful packet delivered to the sink over a sensor network.
As seen in Figure 11, the throughput performance of DelStatic is better than that of NOAH during
congestion. An average throughput value of 8 is recorded during the initial stages of simulation
as compared to 7 of NOAH protocol.

throughput

10
8
6
4

DelStatic

2

NOAH

0
0

5

10

15

time (sec)

Figure 11: Throughput, DelStatic and NoAH

4.5. Part 3: ACCP
The proposed adaptive congestion control seeks the gains of DSR and DelStatic protocol when
used in sensor networks. The metrics of essence are the energy remaining and throughput value
which are essential for a successful data delivery and prolonged sensor life deployment.
We first examine the energy performance of DSR and DelStatic. The energy threshold was set at
500 Joules and the simulation runned for 10 seconds. It was noted as seen in figure 12 that,
DelStatic after controlling congestion had 600J of energy remaining. This is above the set
threshold value. DSR after congestion control had only 100J of energy left.
The sink node of an application using ACCP will use the switching algorithm to trigger DelStatic
anytime energy consumption is below a constant threshold. ACCP protocol is energy efficient.
141


Figure 12: Energy Remaining, ACCP

Figure 13 shows the throughput performance of ACCP. As expected, the figure shows a higher
throughput value for DSR as compared to DelStatic.

throuhgput

10
8
6
4

DelStatic

2

DSR

0
0

5

10

15

time (sec)

Figure 13: Throughput, ACCP

ACCP protocol is recommended for a wireless sensor network application that requires high data
accuracy level and energy efficiency.

5. CONCLUSION
In this paper, we proposed Adaptive Congestion Control protocol which switches between DSR
and DelStatic to control congestion in WSNs based on the nodes remaining energy and priority of
the application. We developed DelStatic by introducing backpressure mechanisms into NOAH.
We chose DSR after our analysis of various on-demand routing protocols showed that DSR has
the highest resource congestion control capability. The simulated results show that ACCP is both
energy and throughput more efficient than existing traffic only or resource only protocols.

REFERENCES
[1]

Wan, C.Y., Eisenman,S.B. and A. T. Campbel, “CODA: Congestion Detection and Avoidance in
Sensor Networks,” in the proceedings of ACM SenSys, Los Angeles, pp. 266279. ACM Press,
New York (2003).

142

[2]

[3]
[4]
[5]
[6]

[7]
[8]
[9]

[10]
[11]

[12]

[13]

[14]
[15]
[16]
[17]
[18]

[19]
[20]

Ee, C.T., Bajcsy, R., “Congestion Control and Fairness for Many-to-One Routing in Sensor
Networks,” in the proceedings of ACMSenSys, Baltimore, pp. 148161. ACM Press, New York
(2004).
Wang, C., Sohraby, K., Li, B., Daneshmand, M., Hu, Y.: “A survery of transport protocols for
wireless sensor networks”, IEEE Network Magazine 20(3), 3440 (2006).
D. B. Johnson and D. A. Maltz, “Dynamic Source Routing in Ad hoc Networks,” Mobile
Computing, ed. T. Imielinski and H. Korth, Kluwer Academic Publishers, 1996, pp. 153-181.
W. Ye, J. Heidemann, and D. Estrin, “An Energy-Efficient Mac Protocol for Wireless Sensor
Networks,” in Proceedings of IEEE INFOCOM’02, June 2002.
Wan, C.Y., Eisenman, S.B. and A. T. Campbel, “CODA: Congestion Detection and Avoidance in
New York (2003).
Wamg. C., Li, B., Sohraby, K., Daneshmand, M., Hu, Y., “A Survey of transport protocols for
wireless sensor networks”, IEEE Network, vol. 20, no. 3, May-June 2006, 34-40.
S. Chen and N. Yan, “Congestion avoidance based on lightweight buffer management in sensor
networks,” in Proceedings of ICPADS, pp. 934 – 946,Sep. 2006.
Wan, C.Y., Eisenman, S.B. and A. T. Campbel “CODA: Congestion Detection and Avoidance in
New York (2003).
Y. Sankarasubramaniam, O. B. Akan, and I.F. Akyidiz, “ESRT: Proceedings of ACM
Mobihoc’03, June 1-3, 2003, Annapolis, USA.
Chonggang Wang, KazemSohraby, Victor Lawrence, Bo Li, Yueming Hu, “Priority-based
Congestion Control in Wireless Sensor Networks”, IEEE International Conference on Sensor
Networks, Ubiquitous, and Trustworthy Computing, Vol 1 (SUTC’06), 2006,pp. 22-31.
Saeed RasouliHeikalabad, Ali Ghaffari, Mir AbolgasemHadian, and HosseinRasouli, “DPCC:
Dynamic Predictive Congestion Control in Wireless,” IJCSI International Journal of Computer
Science Issues, Vol. 8, Issue 1, January 2011, ISSN (Online):1694-0814.R.
Kamal , K.S., Harbhajan, S and, R.B Patel 2010, “A Hop by Hop Congestion Control Protocol to
Mitigate Traffic Contention in Wireless Sensor Networks”, International Journal of Computer
Theory and Engineering, vol. 3, No. 6,pp. 1793-8201.
R. Zheng, J.C. Hou, and L.Sha. Asynchronous Wakeup for Ad Hoc Networks: Theory and
ProtocolDesign.
S. Chen, N. Yang Congestion Avoidance based on Light-Weight Buffer Management in Sensor
Networks, University of Florida, Gainesville, FL 32611,USA.
Hull, B., Jamieson, K., and Balakrishnan, H, Bandwidth management in wireless sensor networks,
Tech. Rep. 909, MIT Laboratory for Computer Science, July2003.
Rangwala, R. Gummadi, R. Govindan, and K. Psounis, “Interference-Aware Fair Rate Control in
Wireless Sensor Networks, SIGCOMM2006.”
Wan, C.Y., Eisenman, S.B., Campbel, A.T.: CODA: Congestion Detection and Avoidance in
Sensor Networks. In: the proceedings of ACM SenSys, Los Angeles, pp. 266279. ACM Press,
New York (2003).
Bertsekas, D. and Gallagher, R. 1991. Data Networks, 2nd ed. Prentice Hall, Englewood Cliffs, NJ,
USA,1991.
Kamal , K.S., Harbhajan, S and, R.B Patel 2010, “A Hop by Hop Congestion Control Protocol to
Mitigate Traffic Contention in Wireless Sensor Networks”, International Journal of Computer
Theory and Engineering, vol. 3, No. 6, pp.1793-8201.

Authors
James D. Gadze received his BSc, MS and PhD degrees in Electrical Engineering from
Kwame Nkrumah University of Science and Technology, Kumasi-Ghana, University of
Tuskegee, Al, US and Florida International University, Miami, US respectively. He is
currently a lecturer in the Electrical & Electronic Engineering department at Kwame Nkrumah
University of Science & Technology in Ghana. His research interest includes, wireless sensor
networks, smart grid security and application of wireless data networks in monitoring and control of critical
infrastructures.

143

DelaliKwasiDake graduated in 2010 with a BSc. Computer Engineering Degree from the
University of Ghana. He also in 2013 had his Master of Philosophy Degree in Computer
Engineering from the Kwame Nkrumah University of Science and Technology. His
entrepreneurial and research ambition makes him a candidate for the new position as a
Lecturer at Lancaster University Ghana. He has detailed research interested in Wireless
Sensor Networks, Data Communication Network and Protocol Design, Artificial Intelligences and
Robotics.
KwasiDiawuo received his BSc, MSc and PhD degrees in Electrical & Electronic Engineering
from the Kwame Nkrumah University of Science and Technology, Kumasi-Ghana; Queen’s
University, Kingston-Canada and University of Newcastle, UK in 1981, 1986 and 1997
respectively. He is currently a Senior Lecturer in the Department of Computer Engineering at
the Kwame Nkrumah University of Science and Technology. His research interests include wireless
networking, Network Protocols, Data communications and Computer Architecture. He is a member of
IEEE.

144

Adaptive congestion control protocol (accp) for wireless sensor networks

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to Adaptive congestion control protocol (accp) for wireless sensor networks (20)

Recently uploaded (20)

Adaptive congestion control protocol (accp) for wireless sensor networks