Downlink massive full dimension-multiple input multiple output downlink beamforming analysis at 3.5 GHz using coordinated ON-OFF switching

International Journal of Electrical and Computer Engineering (IJECE)
Vol. 13, No. 4, August 2023, pp. 4159~4168
ISSN: 2088-8708, DOI: 10.11591/ijece.v13i4.pp4159-4168  4159
Journal homepage: http://ijece.iaescore.com
Downlink massive full dimension-multiple input multiple output
downlink beamforming analysis at 3.5 GHz using coordinated
ON-OFF switching
Chilukuri Nishanthi1
, Nallagarla Ramamurthy2
1
Department of Electronics and Communication Engineering, Jawaharlal Nehru Technological (JNT) University, Anantapur, India
2
Department of Electronics and Communication Engineering, G. Pullaiah College of Engineering and Technology
(Affiliated to JNTUA), Anantapur, India
Article Info ABSTRACT
Article history:
Received Aug 12, 2022
Revised Sep 20, 2022
Accepted Oct 1, 2022
The long-term evolution and advancement (LTE-A) of the 5G wireless
network depends critically on energy consumption. Many existing solutions
focus on limiting power constraints and consequently system coverage. So,
improving the antenna array elements of the base station (BS) can solve this
issue. In this paper, introduce a coordinated ON-OFF switching method in the
massive full dimensional multiple input multiple output (massive-FD-MIMO)
system. It enhances the radiation pattern of the antenna array element by
adjusting the angular power spectra at the BS. By the way, it allows to select
the minimum number of antennas for effective beamforming toward specific
user equipment’s (UEs). In this context, part of antenna element should be
active mode and remining should be sleep mode at the time of signal
beamforming. The multipath spatial profiles are decided the beamforming
frequency band with minimize energy consumption. As part of the method,
we used a conjugated beamforming with power optimization scheme to
determine the individual antenna potential and fading channel condition,
power optimization is performed. This method quality of service, reliability,
energy consumption and data rate can all be evaluated by experimenting with
different-sized antenna arrays such as 16×16, 32×32, 64×64 and 128×128.
Keywords:
Conjugated beamforming
Coordinated ON-OFF
switching
Multipath spatial profile
Interreference contributed
power
UEs specific power allocation
strategy
This is an open access article under the CC BY-SA license.
Corresponding Author:
Chilukuri Nishanthi
Department of Electronics and Communication Engineering, JNT University
Anantapur, Andhra Pradesh, India
Email: chnishanthi@gmail.com
1. INTRODUCTION
Accessibility of user equipment (UE) for data transmission is extended in the 5th generation cellular
network [1]. However, the number of UEs interested in connecting to the 5G network has increased nearly 100
times over 4G, resulting in a huge demand to retain system capacity in terms of data rate (speed >10 Gbps) and
energy efficiency (EE) [2]. In order to meet this requirement, the 5G cellular network must use at least 50% of
total electrical energy for communication [3]. In order to find a new method to optimize power, many
researchers devote more time to it. Over-connected UEs, in particular, should be prioritized in order to
maximize data rates and maintain acceptable quality of service (QoS) levels under heavy data traffic.
Furthermore, massive multiple-input multiple-output (massive MIMO) technology can provide the highest data
rate, low latency and scaled robustness against interference [4]. Because of this, a significant amount of power
is required to maintain the system's capabilities without sacrificing its overall performance [5]. Therefore, to
address this issue an antenna muting principle is developed. This means that only a portion of the antenna

 ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 13, No. 4, August 2023: 4159-4168
4160
element is in use while the rest remains dormant. Energy consumption is therefore reduced, but the method of
turning on and off a minimum number of antenna elements is an open task for any research community. The
frequency bands for 5G communication have been set at 3.4 and 3.8 GHz by IMT-2020. Antenna array design
is heavily influenced by the channel characteristics in this band in terms of both energy consumption and
transmission capability [6]. In this work, a coordinated ON-OFF switching concept in which spatial channel
estimation is used to evaluate the energy efficiency of the massive-full dimension MIMO system under two
important conditions are i) comparing power consumption with respect to the number of UEs for different
antenna array sizes and ii) comparing power consumption along with different percent of traffic conditions
with respect to the number of UEs. This model's antenna array radiation pattern can be improved by adjusting
the base station (BS) power spectrum angularly [7]. In order for beamforming to be effective, a minimum
number of antennas must be selected for desired UEs. Thereby, some antenna elements are activated, whereas
others are deactivated. With the use of multipath spatial profiles, beamforming is performed, which reduces
energy consumption while providing services to the desired UEs. Conjugate beamforming with optimistic power
allocation is used to measure azimuth and elevation power spectra for reliable antenna element selection [8].
The proposed coordinated ON-OFF switching model is improved the energy efficiency of the selected
minimum number of antenna element in the massive-FD-MIMO system. It takes spatial channel estimation in
which analysis the signal interference between UEs through the measured channel coefficients. The
experimental results are conveyed that the power optimization is reached in the IMT-2020 frequency band
around 3.5 GHz. Furthermore, minimum number of active antenna element is measured for effective data
transmission over the available UEs. The following sections frame the remainder of the paper: section 2 reviews
to the recent study, 5G wireless networks in the 3.5 GHz frequency band facing more energy inefficiency than
current 4G networks. System model and its spatial channel estimation, coordinated ON-OFF switching model
and conjugate beamforming with optimistic power allocation scheme are described in section 3. In terms of
QoS, reliability, energy efficiency, data rate, and power consumption, the proposed model's experimental
results are summarized in section 4. Section 5 concludes with a look toward the future.
2. RELATED WORKS
The massive MIMO and spectrum sharing are the important 5G technologies that have had their
energy efficiency examined [7]. The energy and spectral efficiency of D2D devices will be improved by this
new technology. In order to achieve maximum energy efficiency [8], a multi-objective optimization problem
(MOOP) was used to develop a power control system in [9]. For beamforming and spatial multiplexing in a
multi-user large MIMO system, antenna grouping and selection were studied in [10]. Using a low-complexity
binary search technique, the number of transmitting antennas was found. In addition, the impact of combining
two or more energy-saving technologies on network performance, such as UDN with massive MIMO [11], has
been studied. Antenna muting was presented in [12] as a way to improve user experience while also reducing
energy usage. Dynamic antenna muting can be used to transform the MIMO-capable sector coverage mode to
a single transmitting antenna psi-omni mode. Savings of 24.9% are possible with this plan. It was discovered
in [13]–[15] that radio units might be adaptively enabled and deactivated under low load conditions, resulting
in a self-optimizing antenna muting technique. New sleep-based resource allocation strategies have been
proposed in [16], [17] as well. This tactic creates an interference map from which to select the femtocells that
have been turned off. When femtocells are turned off, this map is not updated, which leads to sub-optimal
solutions. Heterogenous networks (HetNets) that use densely deployed femtocells may benefit from an
additional mechanism presented in [18]. The femtocells are put into sleep mode so that they do not interact
with other femtocells. Each femtocell must allow the hybrid access mode in order for UEs in sleep mode to be
connected to another neighboring cell within the cluster using the mechanism. According to [19], an efficient
way to deactivate single-carriers (SCs) that are in undesired interference locations is presented for HetNets,
allowing them to operate more efficiently. In [20], an energy-efficient SC resource allocation technique is
discussed. The algorithm ensures a minimum QoS for multicarrier (MC) UEs while allowing SCs and MCs to
use the same spectrum.
Distributed techniques are investigated [21] for multi-cell beamforming and power allocation without
data sharing between base stations. There is no such thing as a single antenna element in our job. For example,
zero-forcing (ZF) beamforming with a user-selectable selection approach has been studied in terms of data
rates and system performance. ZF beamforming in an frequency division duplex (FDD) system is examined in
[22] to see how it affects the data rate. For massive MIMO systems in uplink transmission, the average data
rate performance of dispersed antennas is studied while all UEs transmit at the same power [23]. When studying
downlink performance of cell-free systems employing time-division duplex (TDD), the maximum-minimum
power allocation technique is applied [24]. There are upper and lower transmission capacity constraints for a
multi-cell massive MIMO system [25]. The asymptotic rate performance of downlink multi-user systems was

Int J Elec & Comp Eng ISSN: 2088-8708 
Downlink massive full dimension-multiple input multiple output downlink … (Chilukuri Nishanthi)
4161
investigated using dispersed antennas and perfect channel state information (CSI). However, the ramifications
of a CSI with fewer access points (APs) and users were addressed [26], [27] for large-scale dispersed antenna
designs. However, in order to deliver consistently satisfactory service to all, or nearly all, customers from a
massive MIMO standpoint, much research is required [27].
3. METHOD
The system paradigm for 5G cellular network beamforming is introduced in this section. Then, to
improve power spectra at the BS and choose the fewest possible antenna array components for effective
beamforming towards the selected UEs. Conjugate beamforming with optimistic power allocation and spatial
channel estimates is an efficient utilization of power based on multipath spatial profiles.
3.1. System model
To illustrate the study, consider a 5G macro cell-centered BS, which can connect to a wide range of
UEs. As shown in Figure 1, the linearly polarized antenna array element is responsible for beamforming to
UEs in both azimuth and elevation directions. The uniform spacing between adjacent elements in the antenna
array is kept at half the carrier wavelength (i.e., 𝑑𝑠𝑝𝑎𝑐𝑒 = 𝜆𝑐𝑎𝑟𝑟𝑖𝑒𝑟 2
⁄ ). The beamforming can be made larger or
smaller depending on the distance, and thus the bandwidth of the transmitting signal is assigned between the
BS and the UEs [28]. The row element in the matrix is in charge of azimuth beamforming, which involves
adjusting the angular direction based on the UE's location. Similarly, the column element is in charge of
elevation beamforming, which involves adjusting the angular direction based on the UE's location. The linearly
polarized antenna array is appropriate for generating multiple beams with respect to different UEs distributed
within the macro cell [29]. In this study, focus on minimum number of active antenna elements selection on
different antenna array sizes includes 16×16, 32×32, 64×64 and 128×128 to achieve a reasonable data rate using
conjugate beamforming with optimistic power allocation to respective UEs for beamforming from BS [30].
Figure 1. System model of coordinated ON-OFF switching model for single macro cell (3.5 GHz)
3.2. Spatial channel estimation
Multipath spatial (MPS) profile is the foundation of the beamforming. To achieve the lowest energy
utility while still offering grantee services to the intended UEs, beamform direction is assessed in channel
coefficient using (1). The downlink of the m-th
antenna element, where m=1, 2, ..., M, and CIR is given by (1):
ℎ𝑠(𝜏, θ) = 𝐹𝑙𝑜𝑠𝑠 ∗ ∑ 𝛼𝑖𝑒𝑗𝜙𝑖𝑆(𝜃𝑖)𝛿(𝜏 − 𝜏𝑖, 𝜃 − 𝜃𝑖)
𝑝
𝑖=1 (1)
where 𝛼𝑖𝑒𝑗𝜙𝑖 represents the complex gains, 𝜏𝑖 represents the long-time delay, and 𝜃𝑖 adjusts the angular
directivity of the i-th
path. Multi-path fading is frequently caused by shadow effects in signal transmission,
which is denoted as 𝐹𝑙𝑜𝑠𝑠. The term 𝛿(𝜏, 𝜃) indicates that the impulse response of the long-time delay exists in
the time domain at 𝜃 = 00
. The MPS's normalized power is defined as ∑ |𝛼𝑖|2
𝑝
𝑖=1 = 1. Subsequently, 𝑆(𝜃𝑖) is the
angular adjustment of the m-th
antenna element with respect to the first antenna element (reference point), from
which the long-time delay between adjacent UEs of the signal beamforming is computed. It is stated in (2).
𝑆(𝜃𝑖) = 𝑒𝑥𝑝 (𝑗
2𝜋
𝜆𝑐𝑎𝑟𝑟𝑖𝑒𝑟
𝑚𝑑𝑠𝑝𝑎𝑐𝑒𝑐𝑜𝑠(𝜃𝑖)) (2)

 ISSN: 2088-8708
4162
Likewise, all M antenna array elements are combined and represented as a vector form, which is given as (3).
𝑆𝑀
= [1 𝑒𝑥𝑝 (𝑗
2𝜋
𝑚𝑑𝑠𝑝𝑎𝑐𝑒𝑐𝑜𝑠(𝜃𝑖)) , . . 𝑒𝑥𝑝 (𝑗
2𝜋
(𝑀 − 1)𝑑𝑠𝑝𝑎𝑐𝑒𝑐𝑜𝑠(𝜃𝑖))] (3)
Therefore, the angular adjustment of the power spectra at the BS is entirely dependent on CRS in order to
improve the antenna array radiation pattern for the UEs downlink.
𝑃𝑑𝑜𝑤𝑛𝑙𝑖𝑛𝑘(𝜃) = ∑ |𝛼𝑖𝑒𝑗𝜙𝑖𝑆(𝜃𝑖)|
2
= ∑ |𝛼𝑖|2
𝜃𝑖=𝜃
𝜃𝑖=𝜃 (4)
For beamforming to desired UEs, all antenna elements have the same angle of adjustment and power distribution.
3.3. Coordinated ON-OFF switching concept
The macro cell capacity is extended up to maximum limits and utilizes the carrier frequency band
(3.5 G Hz) to specific UEs for entire 5G cellular network. Moreover, the signal interference between UEs is
estimated by (1) and improves the beamforming towards UEs through adjusting the antenna diversity gain
measured from the CIR. In the antenna array, set of elements is kept in active mode (turn ON always) to make
balance at the time of coverage impermeability exists [31]. Remaining, antenna element is keep adjusting the
active (ON) and sleepy (OFF) modes continuously depends upon the MPS profile and data traffic. Suppose, if
the antenna element is turn OFF, then, discovery signal is shared to any possible UEs within the macro cell
region. Periodically, coordinated ON-OFF switching is performed to ensure minimum number of antenna
elements are activated for instantaneous conjugate beamforming with optimistic power allocation without
degrades the grantee services to the desired UEs. Figure 2 shows the massive FD-MIMO transceiver
architecture with coordinated ON-OFF switching antenna array elements.
Figure 2. Massive FD-MIMO transceiver architecture with coordinated ON-OFF switching antenna array
elements
Once, the coordinated ON-OFF switching is applied, then, the change of the mode is purely depending
on 𝛽𝑟,𝑐. Where, r and c represent the antenna matrix and there are pre-defined set of antenna element to be turn
ON always denoted as 𝛽𝑟,𝑐
′
. That implies, if 𝛽𝑟,𝑐≠0 = 1, then, it indicates active mode while, 𝛽𝑟,𝑐≠0 = 0, then,
it indicates sleepy mode. The set of all antenna element is considered to be Z and the available UEs within the
macro cell is denoted as 𝑈𝑀. The signal-to-interference noise ratio (SINR) for all available UEs associated to
M antenna element is calculated as (5) [14].
𝑆𝐼𝑁𝑅𝑝,𝑀 =
𝛽𝑀𝑃𝑀𝐺𝑝,𝑀
∑ 𝛽𝑍𝑃𝑍𝐺𝑝,𝑍
𝑍∈𝑀 +𝑁0𝐵𝑊𝑝
(5)
where, 𝑃𝑀 and 𝑃𝑍 denoted as transmitting power of p-th
UEs and total power serving from all antenna element
respectively. The channel gain is achieved between UEs and BS at the macro cell which is denoted as 𝐺𝑝,𝑀.
The noise power spectral density is defined by 𝑁0, and bandwidth of the p-th
UEs is given by 𝐵𝑊
𝑝. To specific
UEs is required maximum data rate for certain instance in that point frequency resource allocation is properly
assigned to the desired UEs based on the conjugated beamforming along with power optimal. By using
Shannon’s equation as mentioned in [19] for data rate of UEs at the p-th
path is given by (6).

4163
𝑅𝑝,𝑀 = 𝐵𝑊
𝑝𝑙𝑜𝑔2(1 + 𝑆𝐼𝑁𝑅𝑝,𝑀) (6)
On the other hand, the minimum number of data rate is achieved by the p-th
UEs from the signal allocated
beamforming is mentioned as (7).
𝑅𝑝,𝑀
𝑚𝑖𝑛
= 𝑅𝑏𝑙𝑜𝑐𝑘 ∗ 𝑅𝑝,𝑀 (7)
where 𝑅𝑏𝑙𝑜𝑐𝑘 be the reserved resource block to ensure the minimum achievable data rate of any possible UEs
within macro cell.
The conjugated beamforming with power optimization is reduced the interference of the UEs and
consequently increases the data rate. According to the proposed model, minimum number of antenna element
is selected and activated to generate beamforming to its respective UEs. Thus, provides only reasonable amount
of power is used to operate the selected antenna element. Therefore, the total power distribution of the entire
antenna array element is defined as (8).
𝑃𝑡𝑜𝑡
𝑎
= 𝑃𝑂𝑁
𝑎
+ 𝐿𝑎𝑃𝑡𝑥
𝑎
(8)
where 𝑃𝑂𝑁
𝑎
be the power related to pre-defined antenna element turn ON always, 𝑃𝑡𝑥
𝑎
is the power consumed
by the antenna element at the time of beamforming to the respective UEs, and 𝐿𝑎 is the transmission loss due
to signal fading. The antenna array element those who are in turn OFF mode, consume some amount of power
due to signal discovery to the any possible UEs. It is denoted as 𝑃𝑂𝐹𝐹
𝑎
= 𝜓𝑃𝑂𝑁
𝑎
. Where, 𝜓 gives the inbuilt
power leakage existing during ON-OFF changes. Now, the total power consumption of the antenna element
can be calculated [20] using (9) given as (9).
𝑃𝑡𝑜𝑡
𝑎
= 𝜓𝑃𝑂𝑁
𝑎
+ (1 − 𝜓)𝛽𝑟,𝑐𝑃𝑂𝑁
𝑎
+ 𝐿𝑎𝑃𝑡𝑥
𝑎
(9)
In this work, the coordinated ON-OFF switching model is minimized power consumption of the linearly
polarized antenna array element based on the MPS profile. However, the interference contribution power (ICP)
is often caused by the active antenna element is calculated [14] with respect to desired UEs given as (10).
𝜒𝑃𝑡𝑥
𝑎 =
∑ 𝑃𝑀𝐺𝑝,𝑖
𝑖∉𝑈𝑀
𝜉 ∑ 𝑃𝑀𝐺𝑝,𝑗
𝑗∉𝑈𝑀
(10)
Because of the continuous turn ON-OFF of the antenna element is included the correction factor
𝜉 = 0.5 𝑙𝑛(𝑀) to compensate the handoffs situation. By increasing the 𝜉 factor, may reduce the ICP at the time
of launch of the beamforming to the UEs. The pseudo code of the coordinated ON-OFF switching model is
described as follows:
Coordinated ON-OFF switching algorithm
Input: No. of UEs within macro cell ‘M’, Size of the antenna array ‘𝑄’, Load threshold 𝐿𝑡ℎ,
Output: Minimum number of active antenna element selection ‘𝑍𝑚𝑖𝑛’
/* Selection based on the MPS profile and ICP */
1. Begin
2. For i=1 to M do
3. If M=0; then
4. 𝑍𝑚𝑖𝑛 → 𝑍𝑂𝐹𝐹 /* To ensure all set of antenna element equal to zero */
5. Else if𝐿𝑎𝑐𝑡𝑢𝑎𝑙 ≥ 𝐿𝑡ℎ, then
6. 𝑍𝑚𝑖𝑛 → 𝑍𝑂𝑁 = 𝑄 2
⁄ ;
7. Else if𝑃𝑡𝑜𝑡
𝑎
< 𝑃𝑡𝑥
𝑎
8. 𝑍𝑚𝑖𝑛 → 𝑍𝑂𝑁 = 𝑄 4
⁄ ;
9. Else
/* Update the antenna element active mode */
10. Find ICP for interchange the active mode between antenna elements using Eq. (10)
11. For i=1 to 𝑄 do
12. If𝜒𝑃𝑡𝑥
𝑎 > 𝜒′
/* Where, 𝜒′
be the ICP threshold limit */
13. 𝑄𝑠𝑒𝑙𝑒𝑐𝑡 → 𝑍𝑜𝑓𝑓
14. Else
15. 𝑄𝑠𝑒𝑙𝑒𝑐𝑡 → 𝑍𝑂𝑁
16. Update𝜒′
17. End if
18. End for
19. End if
20. End for

 ISSN: 2088-8708
4164
3.4. Conjugate beamforming with optimistic power allocation
In this sub-section, discuss about the signal reception by conjugate beamforming with optimistic
power allocation scheme. That means, based on individual antenna potential and channel condition, power
optimization is performed to obtain target data rate with respect to corresponding UEs. There is separate SINR
expression is derived (5) to show the downlink data rate of say M number of UEs. Here, conjugate beamforming
with optimistic power allocation scheme is introduced which is modified version of the specific power
adjustment (SPA) strategy.
a) The number of antenna element in the array is denoted as Q. The beamforming signal gets attenuated due
to fading and thereby, channel coefficient is adjusted which is computed by (1);
b) Each antenna distributes the discovery signal within macro cell and its respective pilot sequence are
𝜑𝑖, 𝑖 = 1,2, … 𝑀;
c) All Q’s estimates the channel coefficient instantaneously, which is given as ℎ𝑠
′ (𝜏, θ);
d) The conjugate beamforming with optimistic power allocation, for the m-th
antenna element's transmitting
signal.
𝑋𝑀 = √𝑃𝑀 ∑ √𝜂𝑝,𝑀
𝑀
𝑚=1 ℎ𝑠
′ (𝜏, θ)𝐷𝑚 (11)
where, 𝐷𝑚 is the data intended to UEs, and 𝜂𝑝,𝑀be the power coefficient used by p-th path of the m-th UEs.
Therefore, the received signal is computed in (12) for the m-th
UEs is given by (12),
𝑌𝑘 = ∑ ℎ𝑠(𝜏, θ)𝑋𝑀 + 𝑊𝑘
𝑀
𝑚=1 (12)
where, 𝑊𝑘~ℕ(0,1) is the additive gaussian noise and ‘k’ represents the only the statistical estimation of
the channel coefficient [7].
4. RESULTS AND DISCUSSION
In this section, the evaluated the performance of the proposed model in different streams includes
i) with use of all set of antenna element is active; ii) a half of the antenna element turn into active based on
MPS profile, and iii) update the turn ON-OFF condition remaining inactive antenna element based on the ICP
threshold. The simulation purpose, the coordinated ON-OFF switching is performed on basis of the ICP and
MPS profile by considering the random selection of antenna element say 50%, 25% of antenna element to be
active (turn ON) along with pre-defined set of existing active antenna element. Moreover, the random induction
of data traffic with similar distribution percentage like 50% and 25% respectively. The actual target is to compute
minimum selection of antenna element to be active for fulfilling the demand request send from UEs in terms of
data rate and energy efficiency. Table 1 shows the simulation parameters which is considered as follows.
Table 1. Simulation parameters
Parameters Values
Transmission power of the macro cell At center 20 dBm and edge 22 dBm
Frequency band 3.5 GHz
Size of the antenna array 16x16, 32x32, 64x64 and 128x128
Number of the UEs 20-120
Baseline power 6.5 Watts
Power spectral density for noise -180 dBm/Hz
Fading loss 𝐹𝑙𝑜𝑠𝑠 = 140 + 35.4𝑙𝑜𝑔10(𝑑), where d is the distance between BS and UEs
in Kms
Mean AoD in the elevation 90°
Mean AoD in the azimuth 0°
Horizontal inter-element spacing, dH 0.5 λ
Vertical inter-element spacing, dV 0.8 λ
Antenna element gain, Gmax; E 8 dBi
Antenna gain, Gmax 15 dBi
Vertical HPBW, 3 dB 15°
Horizontal HPBW, 3 dB 70°
Number of antenna ports in the horizontal direction 4
Number of antenna ports in the vertical direction 8
Figure 3 shows the simulation results of the coordinated ON-OFF switching method under different
random traffic load percentage profile. Figure 3(a) gives the maximum achievable data rate in the consolidated

4165
form of the different antenna array size. It is clearly shown that UEs received the data rate is much higher when
compared to 50% of random traffic load. Reason behind that, the coordinated ON-OFF switching is
significantly influencing the antenna array position and selection numbers which means for lower size antenna
array, say 16×16 cases, for continuous ON-OFF switching, only minimum number of antenna element is
selected for signal beamforming irrespective of the distance. Thereby, there is not much variation when
different percentage of the random traffic load is occurred. If the antenna array size increases, then selection
of antenna element is varying with respect to UEs distance and therefore, probability of occurring signal fading
is more that consumes large amount of the transmission power. It is clearly shown in the Figure 3(b). However,
the antenna array is uniformly spaced between adjacent elements which can adjust the beamforming either
smaller or larger depending on the distance, and thus the bandwidth of the transmitting signal is properly
exchanged between the BS and the UEs.
Figure 4 shows the energy efficiency of the of the coordinated ON-OFF switching method under
varying antenna array size with two different random traffic load (50% and 25%) respectively. According to
continuous ON-OFF switching reduces the number of antenna element which is not required for that instant
signal beamforming to the targeted UEs. However, for internal energy consumption difference between active
and sleep mode antenna element would be considered for performance evaluation. As long as antenna selection
increases, then, energy consumption is more for that energy efficiency is start decreases. Moreover, traffic load
is another distributing factor further consumes energy without need of proper signal validation. Whenever,
higher ratio of random traffic load occurs then it affects the antenna ON-OFF switching which leads to occupy
a greater number of antenna elements for unnecessary signal beamforming. It is clearly shown in Figures 4(a)
and 4(b) respectively. Meanwhile, specific UEs is required maximum data rate for certain instance in that point
frequency resource allocation is properly assigned to the desired UEs based on the conjugated beamforming
along with power optimal. Figure 4 shows the number of antenna element is selected to turn ON-OFF for
making effective signal beam forming. The random traffic load is absent, then the smooth way of antenna
selection is taken and almost the antenna count is maintained constant. It is evidently shown in Figure 5(a).
that means, less than half of the antennas are selected for signal beamforming and if the random traffic load is
interrupted then, some adjustment is carried in the antenna element selection. That implies, last quart half of
the antenna is kept turn OFF position as long as traffic load is reduced. Figure 5(b) stated that the middle of
the time, antenna element gets selected more than the usual antenna element count. it happened due to different
% of the random traffic load occurrence. However, the beamforming may be larger or smaller, allowing the
bandwidth of the transmitting signal to be assigned between the BS and the UEs. Using the UE's location, the
row element of the matrix is responsible for azimuth beamforming. Additionally, the column element is
responsible for adjusting the angle of elevation beamforming based on the user's position. Antenna arrays with
linear polarization can be used to generate multiple beams for different UEs in a macro cell. Therefore,
minimize the energy utility as well as fulfill the grantee services to the desired UEs as shown in Figure 6.
Table 2 shows the comparative analysis of energy efficiency under different % of random traffic load.
(a) (b)
Figure 3. Simulation results under different % of random occurrence of the traffic load: (a) achievable data
rate (bits/s/Hz), and (b) power efficiency (bps/watts)

 ISSN: 2088-8708
4166
(a) (b)
Figure 4. Energy efficiency vs number of antenna elements turn ON: (a) under 50% of random traffic load,
and (b) under 25% of random traffic load
(a) (b)
Figure 5. Number of antenna element selection vs time: (a) without the traffic load condition, and
(b) combined 50% and 25% of traffic load condition
Figure 6. Grantee services to the desired UEs

4167
Table 2. Comparative analysis of energy efficiency under different % of random traffic load
Condition
Macro cell (3.5GHz) at center 20 dBm and edge 22 dBm
Q=16x16, 𝑍𝑚𝑖𝑛 = 12 Q=32x32, 𝑍𝑚𝑖𝑛 = 37 Q=64x64, 𝑍𝑚𝑖𝑛 = 74 Q=128x128, 𝑍𝑚𝑖𝑛 = 88
Without traffic load 5.28 4.96 4.23 3.05
50% Random traffic load 2.38 2.56 2.43 2.27
25% Random traffic load 2.91 2.48 2.83 2.74
According to [28], the number of antenna elements is determined by the size of the antenna array. The
turn-on of the selected antenna elements for successful beamforming is solely determined by the MPS profile
and ICP threshold. That is, the smallest number of antenna elements is chosen, such as 12, 37, 74, and 88 for
16×16, 32×32, 64×64 and 128×128 correspondingly. In the case of lower order antenna arrays, such as 16×16,
the possibility of surpassing the pre-defined threshold level is minimized. Even with erratic data traffic, signal
beamforming with a small number of antenna elements is steady. When the size of the antenna array grows,
there is a greater probability of reducing the sustainability of the minimal number of antenna components that
are activated. However, the suggested approach keeps the minimum number of antenna elements operational
even less than two distinct percentages of random traffic loads.
5. CONCLUSION
In this study, we introduced the Massive-FD-MIMO system to a concept of coordinated ON-OFF
switching. Adjust the antenna array radiation pattern and pick the minimum number of antennas possible for
successful beamforming targeting UEs. The beamforming operation was lowered in energy consumption and
grantee services were supplied to the intended UEs because of the multipath spatial profiles and ICP. As part
of the test, we used a conjugated beamforming with power optimization scheme to determine the individual
antenna potential and fading channel conditions, power optimization is performed, and the minimum number
of antennas turned ON and OFF to obtain achievable data rates with respect to corresponding UEs is ensured.
Experimenting with different-sized antenna arrays, such as 16×16, 32×32, 64×64 and 128×128 can be
evaluated the model's quality of service, reliability, energy consumption, and data rate.
REFERENCES
[1] H. Fourati, R. Maaloul, and L. Chaari, “A survey of 5G network systems: challenges and machine learning approaches,”
International Journal of Machine Learning and Cybernetics, vol. 12, no. 2, pp. 385–431, Feb. 2021, doi: 10.1007/s13042-020-
01178-4.
[2] M. De Ree, G. Mantas, A. Radwan, S. Mumtaz, J. Rodriguez, and I. E. Otung, “Key management for beyond 5G mobile small cells:
A survey,” IEEE Access, vol. 7, pp. 59200–59236, 2019, doi: 10.1109/ACCESS.2019.2914359.
[3] A. Andrae and T. Edler, “On global electricity usage of communication technology: Trends to 2030,” Challenges, vol. 6, no. 1,
pp. 117–157, Apr. 2015, doi: 10.3390/challe6010117.
[4] L. Lu, G. Y. Li, A. L. Swindlehurst, A. Ashikhmin, and R. Zhang, “An overview of massive MIMO: Benefits and challenges,”
IEEE Journal of Selected Topics in Signal Processing, vol. 8, no. 5, pp. 742–758, Oct. 2014, doi: 10.1109/JSTSP.2014.2317671.
[5] X. Zhou, B. Bai, and W. Chen, “Invited paper: Antenna selection in energy efficient MIMO systems: A survey,” China
Communications, vol. 12, no. 9, pp. 162–173, 2015, doi: 10.1109/CC.2015.7275254.
[6] H. Wang and G. Yang, “Design of 4×4 microstrip Quasi-Yagi beam-steering antenna array operation at 3.5 GHz for future 5G
vehicle applications,” in 2017 International Workshop on Antenna Technology: Small Antennas, Innovative Structures, and
Applications (iWAT), 2017, pp. 331–334, doi: 10.1109/IWAT.2017.7915393.
[7] T. L. Marzetta, “Noncooperative cellular wireless with unlimited numbers of base station antennas,” IEEE Transactions on Wireless
Communications, vol. 9, no. 11, pp. 3590–3600, Nov. 2010, doi: 10.1109/TWC.2010.092810.091092.
[8] L. Huang, G. Zhu, and X. Du, “Cognitive femtocell networks: an opportunistic spectrum access for future indoor wireless coverage,”
IEEE Wireless Communications, vol. 20, no. 2, pp. 44–51, Apr. 2013, doi: 10.1109/MWC.2013.6507393.
[9] D. Della Penda, L. Fu, and M. Johansson, “Energy efficient D2D communications in dynamic TDD systems,” IEEE Transactions
on Communications, vol. 65, no. 3, pp. 1260–1273, Mar. 2017, doi: 10.1109/TCOMM.2016.2616138.
[10] A. Kaushik et al., “Spectrum sharing for 5G wireless systems (spectrum sharing challenge),” in 2015 IEEE International Symposium
on Dynamic Spectrum Access Networks (DySPAN), Sep. 2015, vol. 20, no. 2, pp. 1–2, doi: 10.1109/DySPAN.2015.7343841.
[11] S. Shalmashi, E. Bjornson, M. Kountouris, K. W. Sung, and M. Debbah, “Energy efficiency and sum rate when massive MIMO
meets device-to-device communication,” in 2015 IEEE International Conference on Communication Workshop (ICCW), Jun. 2015,
pp. 627–632, doi: 10.1109/ICCW.2015.7247251.
[12] Z. Al-Husseiny and P. Frenger, “Enhancing LTE energy performance with antenna muting and dynamic psi-omni configuration,”
in 2015 IEEE 81st Vehicular Technology Conference (VTC Spring), May 2015, pp. 1–5, doi: 10.1109/VTCSpring.2015.7145824.
[13] M. Amirijoo, Z. Chai, P. Frenger, B. Olin, and J. Moe, “Self-optimizing antenna muting - energy consumption and user thoughput
analysis,” in 2012 International Symposium on Wireless Communication Systems (ISWCS), Aug. 2012, pp. 46–50, doi:
10.1109/ISWCS.2012.6328327.
[14] B. Shen, Z. Lei, X. Huang, and Q. Chen, “An interference contribution rate based small cells on/off switching algorithm for 5G
dense heterogeneous networks,” IEEE Access, vol. 6, pp. 29757–29769, 2018, doi: 10.1109/ACCESS.2018.2841044.
[15] A. Ebrahim and E. Alsusa, “Interference minimization through sleep mode based resource allocation for future femtocell networks,”
in 2015 IEEE International Conference on Communications (ICC), Jun. 2015, pp. 1679–1684, doi: 10.1109/ICC.2015.7248566.
[16] C. Bouras and G. Diles, “Energy efficiency in sleep mode for 5G femtocells,” in 2017 Wireless Days, Mar. 2017, pp. 143–145, doi:
10.1109/WD.2017.7918130.

 ISSN: 2088-8708
4168
[17] A. Ebrahim and E. Alsusa, “Interference and resource management through sleep mode selection in heterogeneous networks,” IEEE
Transactions on Communications, pp. 1–1, 2016, doi: 10.1109/TCOMM.2016.2623614.
[18] A. Saeed, E. Katranaras, A. Zoha, A. Imran, M. A. Imran, and M. Dianati, “Energy efficient resource allocation for 5G
heterogeneous networks,” in 2015 IEEE 20th International Workshop on Computer Aided Modelling and Design of Communication
Links and Networks (CAMAD), Sep. 2015, vol. 9, pp. 119–123, doi: 10.1109/CAMAD.2015.7390492.
[19] Y. Ni, J. Liang, X. Shi, and D. Ban, “Research on key technology in 5G mobile communication network,” in 2019 International
Conference on Intelligent Transportation, Big Data & Smart City (ICITBS), Jan. 2019, pp. 199–201, doi:
10.1109/ICITBS.2019.00054.
[20] X. Huang, S. Tang, Q. Zheng, D. Zhang, and Q. Chen, “Dynamic femtocell gNB on/off strategies and seamless dual connectivity
in 5G heterogeneous cellular networks,” IEEE Access, vol. 6, pp. 21359–21368, 2018, doi: 10.1109/ACCESS.2018.2796126.
[21] Y. Huang, G. Zheng, M. Bengtsson, K.-K. Wong, L. Yang, and B. Ottersten, “Distributed multicell beamforming with limited
intercell coordination,” IEEE Transactions on Signal Processing, vol. 59, no. 2, pp. 728–738, Feb. 2011, doi:
10.1109/TSP.2010.2089621.
[22] O. Somekh, O. Simeone, Y. Bar-Ness, A. M. Haimovich, and S. Shamai, “Cooperative multicell zero-forcing beamforming in
cellular downlink channels,” IEEE Transactions on Information Theory, vol. 55, no. 7, pp. 3206–3219, Jul. 2009, doi:
10.1109/TIT.2009.2021371.
[23] H. Huh, A. M. Tulino, and G. Caire, “Network MIMO with linear zero-forcing beamforming: Large system analysis, impact of
channel estimation, and reduced-complexity scheduling,” IEEE Transactions on Information Theory, vol. 58, no. 5, pp. 2911–2934,
May 2012, doi: 10.1109/TIT.2011.2178230.
[24] K. T. Truong and R. W. Heath, “The viability of distributed antennas for massive MIMO systems,” in 2013 Asilomar Conference
on Signals, Systems and Computers, Nov. 2013, pp. 1318–1323, doi: 10.1109/ACSSC.2013.6810508.
[25] J. Wang and L. Dai, “Asymptotic rate analysis of downlink multi-user systems with co-located and distributed antennas,” IEEE
Transactions on Wireless Communications, vol. 14, no. 6, pp. 3046–3058, Jun. 2015, doi: 10.1109/TWC.2015.2399921.
[26] H. Yin, D. Gesbert, and L. Cottatellucci, “Dealing with interference in distributed large-scale MIMO systems: A statistical
approach,” IEEE Journal of Selected Topics in Signal Processing, vol. 8, no. 5, pp. 942–953, Oct. 2014, doi:
10.1109/JSTSP.2014.2322583.
[27] H. Yang and T. L. Marzetta, “Capacity performance of multicell large-scale antenna systems,” in 2013 51st Annual Allerton
Conference on Communication, Control, and Computing (Allerton), Oct. 2013, pp. 668–675, doi: 10.1109/Allerton.2013.6736589.
[28] D. Pinchera, M. Migliore, F. Schettino, and G. Panariello, “Antenna arrays for line-of-sight massive MIMO: Half wavelength is not
enough,” Electronics, vol. 6, no. 3, Aug. 2017, doi: 10.3390/electronics6030057.
[29] K. S. Ananda Kumar and R. Balakrishna, “Evaluation of energy consumption using receiver–centric MAC protocol in wireless
sensor networks,” International Journal of Electrical and Computer Engineering (IJECE), vol. 8, no. 1, pp. 87–93, Feb. 2018, doi:
10.11591/ijece.v8i1.pp87-93.
[30] D. S. Sisodia, R. Singhal, and V. Khandal, “A performance review of intra and inter-group MANET routing protocols under varying
speed of nodes,” International Journal of Electrical and Computer Engineering (IJECE), vol. 7, no. 5, pp. 2721–2730, Oct. 2017,
doi: 10.11591/ijece.v7i5.pp2721-2730.
[31] N. Kirubanandasarathy, K. Kottaisamy, and M. Rajmohan, “Study and survey about mixed radix FFT processor in MIMO OFDM,”
International Journal of MC Square Scientific Research, vol. 2, pp. 47–52, 2010.
BIOGRAPHIES OF AUTHORS
Chilukuri Nishanthi received the B.Tech degree in electronics and communication
engineering from Bhoj Reddy Engineering college, Hyderabad, in 2010 also the M.Tech. in in
electronics and communication engineering from Jagruthi Institute of Engineering and
Technology, Ibrahimpatnam, Telangana in 2013 and pursuing Ph.D. degree in in electronics and
communication engineering from Jawaharlal Nehru Technological University Anantapur, AP.
Currently, she is an Associate Professor at the Department of Electronics and Communication
Engineering, Teegala Krishna Reddy Engineering College, Autonomous University. Her
research interests include wireless communication technology, signal processing, 5G
technology, communication systems, digital transformations, MIMO technology, and artificial
intelligence applied power system. Her research interest made her to publish research papers in
international national conferences, also published papers in reputed journals like Elsevier and
springer and few UGC approved journals. She can be contacted at chnishanthi@mail.com.
Nallagarla Ramamurthy received his B.Tech. and M.Tech degrees in 1998 and
2006, respectively, from the Department of Electronics and Communication Engineering, Sri
Venkateshwara University, Ananthapuramu, Andhra Pradesh, and his Ph.D. degree in 2014 from
the Department of Electronics and Communication Engineering, Jawaharlal Nehru
Technological University, Anantapur. He is presently working as Professor in G. Pullaiah
College of Engineering and Technology Affiliated to JNTUA, Ananthapuramu, Andhra Pradesh.
He has 23 years of teaching faculty experience. His general research interests include Digital
Image processing and Signal Processing communications, in these areas, he has published over
60 registered national and international journals. He is one of the reviewers for the IJERT journal
in the year 2017. He can be contacted at email: ramamurthy1006@gmail.com.

Downlink massive full dimension-multiple input multiple output downlink beamforming analysis at 3.5 GHz using coordinated ON-OFF switching

More Related Content

Similar to Downlink massive full dimension-multiple input multiple output downlink beamforming analysis at 3.5 GHz using coordinated ON-OFF switching (20)

More from IJECEIAES (20)

Recently uploaded (20)

Downlink massive full dimension-multiple input multiple output downlink beamforming analysis at 3.5 GHz using coordinated ON-OFF switching