This section presents the performance evaluation for the proposed NOMA-enhanced BackCom, both static and dynamic systems. We also comparatively discuss the perfor-
53
mance of our proposed system with the conventional one reported in [8]. Parameters used for the simulations, unless otherwise noted, are summarized in Table. 3.1. More- over, Table. 3.2 details the data structure of a BN through a record. The Monte-Carlo simulations are implemented to verify the accuracy of analytical derivations presented in the previous sections. Further, the flowchart simulation process of our proposed schemes are shown in Fig. 3.3, Fig. 3.4 and Fig. 3.5. The simulation results have been obtained by averaging over 1000 iteration runs.
Table 3.1: Simulation parameters for NOMA-aided BackCom systems.
Symbol Description Values
B Number of BNs 60
BN data Backscatter node’s data Table. 3.2
M Number of NOMA groups 2, 5
RO Outer radius of coverage zone 100 [m]
RI Inner radius of coverage zone 2 [m]
ξi, i ∈ [1, 5] Power reflection coefficient [0.7, 0.5, 0.3, 0.1, 0.05]
P Reader’s transmitted power 25 [dBm]
No Noise power -90 [dBm]
ρ Path-loss coefficient 2.5
Ts Timeslot duration 1 [second]
γth Channel threshold 0 ÷ 20 [dB]
Number of transmitted bits during a mini-slot [8] 60
Table 3.2: Backscatter node’s data structure.
Symbol Description Unit
X Horizontal coordinates of BN (X-axis) [m]
Y Vertical coordinates of BN (Y-axis) [m]
r Distance between a BN and a reader [m]
Subregion Subregion to which the BN belongs - ξ Power reflection coefficient - Pri Backscatter power from the BN [mW]
Po Interference power from the BN [mW]
SINR Signal-to-Interference-to-Noise Rate -
3.4.1. Number of Successful Backscatter Nodes
We now investigate the normalized number of successful BNs, which is normalized by the total number of BNs. First, we quantitatively highlight the effectiveness of our proposed schemes, i.e., two-node pairing (TNP) and adaptive power reflection coefficient (APRC), in comparison with the conventional one [8] in Fig. 3.6. Different channel thresholds, γth, are taken into account. As expected, our proposed schemes, i.e., both TNS and APRC schemes, outperform the conventional one over a range
Figure 3.3: The flowchart of TNP scheme.
of channel thresholds. This is because these schemes could increase the possibility of successful decoding in NOMA groups by pairing selected BNs, which is not in a random manner (the TNS scheme), or adjusting the power reflection coefficients depending on the channel conditions (the APRC scheme). When comparing the TNS scheme with the APRC scheme, it is observed that the TNS scheme can achieve a better performance than that of the APRC scheme. However, the complexity of the TNS scheme increases when the NOMA group size increase, while the APRC scheme offers a simpler solution for large values of NOMA group size. As a result, it is recommended to use the TNS scheme and the APRC scheme for small (M = 2) and large values (M > 2) of NOMA group size. For example, when γth = 12 dB, the normalized number of successful BNs are 0.98, 0.68, and 0.5 for the TNS, APRC, and conventional schemes, respectively.
Figure 3.7 illustrates the normalized number of successful BNs over a range of chan- nel thresholds in the static NOMA-enhanced BackCom systems using the TNS scheme.
Also, different reader’s transmitter powers, i.e., P = 21 dBm, 23 dBm, 25 dBm, 27
Initialization BNs are sorted according
to their power levels
End Set all other BNs in
far-subregion to transmit in mini-slot
for single-node Near-
subregion is empty?
Yes
No
SINR ? Yes No
Set BNs in a NOMA group to transmits in mini-slot for 2-node NOMA group
Set BN in near- subregion to transmits in mini- slot for single-node
Pair NOMA group Calculate SINR BNs are divided into 2 subregions (Near and Far)
55
Figure 3.4: The flowchart of APRC scheme.
dBm, are investigated. Using this figure, we can decide the power levels corresponding to the channel thresholds to retain a target performance. For instance, to maintain the normalized number of successful BNs of 0.8, the power levels with the channel thresholds are (P , γth) = (21 dBm, 10 dB), (23 dBm, 12 dB), (25 dBm, 14 dB), and (27 dBm, 16 dB). Also, in this figure, the analytical results follow the simulated ones very closely, which confirms the model’s correctness and analysis.
The performance of the APRC scheme in terms of the normalized number of suc- cessful BNs is illustrated in Fig. 3.8. Also, different values of the NOMA group size, i.e., M = 2, 3, 4, 5, are taken into account. As is evident, when the NOMA group size increases, the normalized number of successful BNs decreases. The reason is that a larger NOMA group size reduces the possibility of successful decoding in NOMA
Initialization BNs are sorted according
to their power levels
End Set all other BNs in
remaining subregion to transmit in mini- slot for single-node Any
subregion is empty?
Yes
No
SINR ? Yes No
Set BNs in a NOMA group to transmits in mini-slot for M-node NOMA group
Set BN to transmits in mini-slot for single-node
Pair NOMA group Calculate SINR
i
i Pr 2
Pj No opt M
j i 1
th
i
Adjust i to successfully decode BN BNs are divided into
M subregions
Figure 3.5: The flowchart of DSP scheme.
systems. For example, when the channel threshold of γth = 10 dB, the normalized number of successful BNs for the APRC scheme are 0.1, 0.15, 0.38, and 0.95 for the NOMA group size of M = 5, 4, 3, 2, and 1, respectively.
In Fig. 3.9, we investigate the performance of dynamic NOMA-enhanced BackCom systems using the DSP scheme. Also, different reader’s transmitter powers, i.e., P = 21 dBm, 23 dBm, 25 dBm, 27 dBm, are considered. As seen, the higher values of the normalized number of successful BNs require higher levels of transmitted power. For instance, to maintain the normalized number of successful BNs of 0.8, the transmitted power levels required are 21 dBm, 23 dBm, 25 dBm, and 27 dB, for the channel thresholds γth of 12 dB, 14 dB, 16 dB, and 18 dB, respectively.
B.size() 0 No
End Yes