Abstract
In this paper, we propose two adaptive cooperation protocols for wireless systems. The first protocol chooses between cooperative communications (CC) and non-cooperative communications (NCC), the protocol that offers the highest instantaneous throughput. The best relay is activated only if the instantaneous throughput of relayed link is larger than that of direct link. The second protocol chooses between CC and NCC, the protocol that maximizes the average throughput. The best relay amplifies the received signal only when the average throughput of relayed link is larger than that of direct link. Both protocols include adaptive modulation and coding to increase the throughput by selecting the appropriate modulation and coding scheme. The results are valid for Rayleigh fading channels where relay nodes use amplify and forward relaying.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Adaptive modulation and coding (AMC) adapts the modulation and coding scheme (MCS) to channel propagation conditions [1,2,3,4]. When the instantaneous signal-to-noise ratio (ISNR) is high, we can increase the rate of channel encoder to reduce parity bits as well as increasing the constellation size [5]. In all previous studies [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15], only the MCS is adapted to ISNR. The AMC technique can be improved when combined with Cooperative Diversity (CD) [16,17,18,19,20,21,22]. The ISNR can be high as the best relay with largest ISNR is activated. In [16,17,18,19,20,21,22], AMC is used with cooperative diversity and the relay is always active. The main drawback of this approach is that the throughput is half that of non-cooperative diversity since half the duration of the frame is used by the source and the rest of the frame is used by the relay. In [23, 24], Incremental Relaying (IR) has been implemented with a relay equipped with an energy buffer. IR uses the direct link when its ISNR is larger than threshold T. Otherwise, the relayed link is used since the ISNR of direct link is lower than T. In [25], IR has been extended to energy harvesting systems. IR for non-orthogonal multiple access (NOMA) has been suggested in [26]. IR can be used in primary or secondary networks of cognitive radio systems [27].
In this paper, we compare two adaptive cooperation (AC) protocols with AMC based on ISNR or Average SNR (ASNR). We select the best MCS and best transmission protocol between non-cooperative and cooperative communications. The main contributions of our paper are:
-
We suggest an AC protocol with AMC that offers higher throughput than [16,17,18,19,20,21,22] where the relays are always active. In this paper, we activate the relays only when required, i.e., when the corresponding throughput is larger than that of direct link (NCC: Non-Cooperative Diversity).
-
Two adaptive cooperation (AC) protocols are suggested in order to maximize the average or instantaneous throughput using average or instantaneous SNR (ASNR or ISNR).
-
The proposed AC protocol offers larger throughput than Incremental Relaying (IR) [23,24,25,26,27]. AC offers 1.2–1.3 dB gain with respect to IR.
Section 2 describes the system model. Section 3 deals with AC using ISNR to maximize the instantaneous throughput. Section 4 deals with AC using ASNR. Section 5 provides some theoretical and simulation results. Section 6 studies the complexity of AC. Last section concludes the paper.
2 System model
The system model is shown in Fig. 1. There are a source S, K relays \(R_k\) and a destination D. In the proposed AC protocol, we begin by selecting the best relay that offers the highest end-to-end SNR at destination D. We choose the best modulation and coding scheme (MCS), and we compute the average or instantaneous throughput.
3 Adaptive cooperation using ISNR
3.1 Throughput analysis of non-cooperative communications
The instantaneous throughput of NCC is written as
where M is the size of modulation, L is packet length in symbols, the number of bits per packets is \(L\times \log _2(M)\), \(\gamma _{\mathrm{SD}}\) is the ISNR between the source S and destination D, \(\hbox {PEP}(\gamma _{\mathrm{SD}})\) is the packet error probability (PEP) at D for ISNR equal to \(\gamma _{\mathrm{SD}}\):
\(\hbox {SEP}(\gamma _{\mathrm{SD}})\) is the Symbol Error Probability at D for ISNR equal to \(\gamma _{\mathrm{SD}}\).
For M-QAM modulation, the SEP is expressed as [28]
For M-ASK modulation, the SEP is written as [28]
For M-PSK modulation, we have [28]
For Coded NCC with an \(R_{\mathrm{c}}\) rate convolutional encoder, the ITHR is equal to
where the number of information bits per packet is \(R_{\mathrm{c}}\times L \times \log _2(M)\) and \(\hbox {EEP}(\gamma _{\mathrm{SD}})\) is the error event probability (EEP) of Viterbi channel decoder written as [28]
\(d_f\) and \(a_d\) are the free distance and distance spectra [28].
The Average Throughput of NCC is equal to
where \(f_{\varGamma _{\mathrm{SD}}}(\gamma _{\mathrm{SD}})\) is the Probability Density Function (PDF) of SNR and \(\overline{\varGamma _{\mathrm{SD}}}\) is the ASNR between S and D,
\(\hbox {ITHR}_{\mathrm{NCC}}^{\mathrm{AMC}}(\gamma _{\mathrm{SD}})\) is the ITHR when AMC is used which is the maximum of throughput of the different MCS:
For Rayleigh fading channels, the ISNR follows an exponential distribution written as [28]
Four modulation and coding schemes (MCS) are studied in this paper, but our results are valid for other MCS: MCS1 (\(M=4\), \(R_{\mathrm{c}}=0.5\)), MCS2 (\(M=4\), \(R_{\mathrm{c}}=1\)), MCS3 (\(M=16\), \(R_{\mathrm{c}}=1\)), MCS4 (\(M=64\), \(R_{\mathrm{c}}=1\)).
3.2 Throughput analysis of cooperative communications
The ITHR of cooperative communications (CC) is written as
There is coefficient \(\frac{1}{2}\) since half the frame is dedicated for transmission by S and half duration for transmission by relay R, \(\gamma _{{SRD}}\) is the ISNR between the source S, Relay R and destination D.
For Coded CC with an \(R_{\mathrm{c}}\) rate convolutional encoder, the ITHR is equal to
The throughput of CC is expressed as
where \(f_{\varGamma _{{SRD}}}(\gamma _{{SRD}})\) is the PDF of SNR and \(\overline{\varGamma _{{SRD}}}\) is the ASNR of link \(S{-}R{-}D\), \(\hbox {ITHR}_{\mathrm{CC}}^{\mathrm{AMC}}(\gamma _{{SRD}})\) is the ITHR when AMC is used which is the maximum of throughput of the different MCS:
The ISNR of amplify and forward (AF) relaying is defined as [29]
where \(\gamma _{SR}\) (respectively \(\gamma _{RD}\)) is the ISNR between the source S and relay node R (respectively between R and destination D). For Rayleigh fading channels, the PDF of ISNR is given by [30]
where \(\gamma _{{SRD}}>0\), \(\overline{\varGamma }_{SR}\) (respectively \(\overline{\varGamma }_{RD})\) are the ASNR between S and R (respectively R and D).
When there are K relays, we can select the relay \(R_{\mathrm{sel}}\) that offers the highest end-to-end SNR:
K is the number of relays.
Assuming that the SNR due to different relays are independent, the Cumulative Distribution Function (CDF) of the SNR \(\varGamma _{SR_{\mathrm{sel}}D}\) is expressed as
The PDF of \(\varGamma _{SR_{\mathrm{sel}}D}\) is equal to
where \(f_{\varGamma _{SR_{n}D}}(x)\) is expressed as (16) and \(F_{\varGamma _{SR_{k}D}}(x)\) is given by [25]
In the presence of many relays, the average throughput is given in (13) where we have to use the PDF of SNR given in (19).
3.3 Adaptive cooperation using ISNR
The proposed adaptive cooperation (AC) selects between NCC and CC the protocol that has the highest instantaneous throughput
The throughput of AC is written as
AC maximizing the instantaneous throughput is implemented as follows:
-
Compare the ISNR of direct link to thresholds \(T_1\), \(T_2\) and \(T_3\) of Fig. 2 and select the appropriate MCS. If \(ISNR< T_1\), QPSK with half rate channel coding is used. If \(T_1<ISNR\le T_2\), uncoded QPSK is used. If \(T_2<ISNR\le T_3\), uncoded 16QAM is used. If \(\hbox {ISNR}>T_3\), uncoded 64QAM is used. Compute the instantaneous throughput of direct link.
-
Compare the ISNR of relaying links and select the relay with largest instantaneous throughput. For the selected relay, compare the corresponding ISNR to thresholds \(T_1\), \(T_2\) and \(T_3\) of Fig. 2 and select the appropriate MCS. Compute the instantaneous throughput of relayed link.
-
Compare the instantaneous throughput of NCC and NCC and use the protocol with highest throughput.
4 Adaptive cooperation using ASNR
In this section, we select the MCS and cooperation/non-cooperation protocol using ASNR per bit expressed as \(\frac{E_{\mathrm{b}}}{N_0}\).
The throughput of NCC is expressed as (8).
When AMC using the ASNR is employed, the throughput of NCC becomes
The throughput of cooperative communications (CC) is expressed similarly
where the instantaneous throughput \(\hbox {ITHR}_{\mathrm{CC}}^{R_{\mathrm{c}},M}(\gamma _{{SRD}})\) is given in (11) and (12) for uncoded and coded communications.
When AMC using the ASNR is employed, the average throughput of CC becomes
The adaptive cooperation (AC) protocol selects the best protocol between CC and NCC and the best MCS using ASNR
AC maximizing the average throughput is implemented as follows:
-
Compare the ASNR of direct link to thresholds \(A_1\), \(A_2\) and \(A_3\) and select the appropriate MCS. The values of thresholds are \(A_1=6.3 \,\hbox {dB}\), \(A_2=9.2 \,\hbox {dB}\) and \(S_3=16.2 \,\hbox {dB}\) and allow to select the appropriate MCS maximizing the average throughput. If \(\hbox {ASNR}\le 6.3\,\hbox {dB}\), QPSK with half rate channel coding is used. If \(6.3 \,\mathrm{\,dB}<\hbox {ASNR}\le 9.2 \,\hbox {dB}\), uncoded QPSK is used. If \(9.2\,\hbox {dB}<\hbox {ASNR}\le 16.2 \,\hbox {dB}\), uncoded 16QAM is used. If \(ASNR>16.2 \,\hbox {dB}\), uncoded 64QAM is used. Compute the average throughput of direct link.
-
Compare the ASNR of relaying links and select the relay with largest average throughput. For the selected relay, compare the corresponding ASNR to thresholds \(A_1\), \(A_2\) and \(A_3\) and select the appropriate MCS. Compute the average throughput of relayed link.
-
Select the best protocol between CC and NCC in order to maximize the throughput.
5 Numerical results
Path loss exponent is equal to three. The distance on direct link is one. The distance between source and relays is \(d_{SR_k}\)=0.2,0.3,0.4,0.5. The distance between relays and destination is \(d_{R_kD}=1-d_{SR_k}\). These are normalized distance \(d_{SR_k}=\frac{d_{SR_k,\mathrm{effective}}}{d_0}\) and \(d_{R_kD}=\frac{d_{R_kD,\mathrm{effective}}}{d_0}\) where \(d_0\) is a reference distance in meters and \(d_{SR_k,\mathrm{effective}}\) and \(d_{R_kD,\mathrm{effective}}\) are the distances in meters between S and \(R_k\) and \(R_k\) and D. Packet length is \(L\times \log _2(M)=600\) bits.
Figure 2 shows the instantaneous throughput of NCC with respect to ISNR. These curves correspond to the studied MCS. We following strategy should be used.
-
If the \(\hbox {ISNR} < T_1=6.3 \,\hbox {dB}\), the best MCS is MCS1: \(R_{\mathrm{c}}=0.5\) with QPSK modulation.
-
If \(T_1<\hbox {ISNR}< T_2=10.7 \,\hbox {dB}\), the best MCS is MCS2: \(R_{\mathrm{c}}=1\) with QPSK modulation.
-
If \(T_2<\hbox {ISNR}<T_3=15.7 \,\hbox {dB}\), the best MCS is MCS3: \(R_{\mathrm{c}}=1\) with 16-QAM.
-
If ISNR\(>T_3\), the best MCS is MCS4: \(R_{\mathrm{c}}=1\) with 64-QAM.
As shown in Fig. 2, the thresholds \(T_i\) are the abscissae of the intersections between the throughput of the studied 4 MCS.
Figure 3 shows the throughput of adaptive cooperation (AC) using ISNR in the absence of path loss and a single relay. The red curve corresponds to AC with AMC using ISNR. The blue curve corresponds to AC with 64-QAM. The black curve corresponds to Non-Cooperative Communications NCC (direct link) with AMC using ISNR. The magenta curve corresponds to cooperative communications (CC) with AMC using ISNR. We observe that the proposed AC offers the highest throughput. For a throughput of 4 bit/s/Hz, the proposed AC offers 2 dB gain with respect to NCC with AMC. Also, AC has a higher throughput than CC with AMC as studied in [16,17,18,19,20,21,22].
Figure 4 shows the throughput of AC and CC with AMC using ISNR in the presence of 1, 2, 3 and 4 relays. The distance between all nodes is equal to one. The relay with largest ISNR is selected. We observe that AC offers a higher throughput than CC even in the presence of multiple relays.
Figure 5 shows the throughput of AC using ASNR in the presence of a single relay. We observe that AC offers a higher throughput than CC and NCC protocols. In fact, the proposed AC chooses the best protocol that ensures the highest throughput.
Figure 6 shows the throughput of AC with AMC using the ASNR and ISNR. We observe that AC based on ISNR offers the highest throughput. AC using ISNR offers 10.5 dB gains with respect to AC using ASNR for a throughput of 4bit/s/Hz. In fact, AC with ISNR selects the appropriate MCS and CC/NCC protocol based on instantaneous channel conditions.
Figure 7 compares the throughput of proposed adaptive cooperation (AC) protocol using ISNR with respect to Incremental Relaying (IR) [23,24,25,26,27] in the presence of a single relay. Both AC and IR use AMC and choose the best MCS. IR use the direct link when its SNR is larger than threshold T. Otherwise, the relayed link is used since the direct link has a low SNR. We observe that proposed AC offers higher throughput than IR for all values of \(T=1,5,10\). A throughput of 4 bit/s/Hz is reached for \(E_{\mathrm{b}}/N_0=8.4\) dB when AC is used. The throughput of IR with \(T=5\) is equal to 4 bit/s/Hz for \(E_{\mathrm{b}}/N_0=9.6\) dB. A throughput of 3 bit/s/Hz is reached for \(E_{\mathrm{b}}/N_0\) equal to 6 dB and 7.3 dB, respectively, for AC and IR with \(T=5\). Therefore, AC offers 1.2 dB (respectively 1.3 dB) gain with respect to IR for a throughput of 4 bit/s/Hz (respectively 3 bit/s/Hz). AC chooses always the best transmission protocol. However, IR is suboptimal since it chooses the relayed link when the SNR of direct link is less than T. When the SNR of direct link is less than T, it is possible that the throughput of relayed link is lower throughput than that of direct link.
6 Complexity analysis
AC maximizing the instantaneous throughput requires to measure the ISNR of direct and relayed links. This estimation is used to determine the best protocol between CC and NCC. Also, the ISNR is used to determine the best MCS. AC maximizing the average throughput requires to measure the average SNR of direct and relayed links. NCC (respectively CC) requires to measure the ISNR of direct link (respectively relayed link) to choose the best MCS. IR requires also to measure both ISNR of direct and relayed links since it is used to choose the best MCS and to know whether the relay will be idle or activated. Therefore, AC has a similar complexity as IR. However, the proposed AC offers 1.2–1.3 dB gain with respect to IR. NCC selects the appropriate MCS in 1.11 ms. CC selects the appropriate MCS and relay in 1.13ms. AC selects the appropriate protocol, relay and MCS in 1.17 ms.
7 Conclusions
In this paper, we have suggested a new AC protocol that chooses the best protocol between CC and NCC in order to maximize the throughput. We use ISNR or ASNR to choose between NCC and CC. AC uses the relayed link only when its throughput is larger than the direct link. The proposed adaptive cooperation protocol has been extended to include adaptive modulation and coding (AMC) and relay selection and reaches higher throughput than non-cooperative protocol and cooperative protocol with AMC. For a throughput of 4 bit/s/Hz, the proposed AC offers 2 dB gain with respect to NCC with AMC. AC using ISNR offers better performance than AC using ASNR. The proposed AC protocol offers 1.3 dB gain with respect to incremental relaying for a throughput of 3 bit/s/Hz.
References
Liu, Q., Zhou, S., Giannakis, G.B.: Cross-layer combining of adaptive modulation and coding with truncated ARQ over wireless links. IEEE Trans. Wirel. Commun. 3(5), 1746–1755 (2004)
Goldsmith, A.J., Chua, S.-G.: Adaptive coded modulation for fading channels. IEEE Trans. Commun. 46(5), 595–602 (1998)
Liu, Q., Zhou, S., Giannakis, G.B.: Queuing with adaptive modulation and coding over wireless links: cross-layer analysis and design. IEEE Trans. Wirel. Commun. 4(3), 1142–1153 (2005)
Catreux, S., Erceg, V., Gesbert, D., Heath, R.W.: Adaptive modulation and MIMO coding for broadband wireless data networks. IEEE Commun. Mag. 40(6), 108–115 (2002)
Hole, K.J., Holm, H., Oien, G.E.: Adaptive multidimensional coded modulation over flat fading channels. IEEE J. Sel. Areas Commun. 18(7), 1153–1158 (2000)
Oien, G.E., Holm, H., Hole, K.J.: Impact of channel prediction on adaptive coded modulation performance in Rayleigh fading. IEEE Trans. Veh. Technol. 53(3), 758–769 (2004)
Wang, X., Liu, Q., Giannakis, G.B.: Analyzing and optimizing adaptive modulation coding jointly with ARQ for QoS-guaranteed traffic. IEEE Trans. Veh. Technol. 56(2), 710–720 (2007)
Koike-Akino, T., Kojima, K., Millar, D., Parsons, K., Yoshida, T., Sugihara, T.: Pareto optimization of adaptive modulation and coding set in nonlinear fiber-optic systems. J. Lightw. Technol. PP(99), 1 (2016)
López-Benítez, M.: Throughput performance models for adaptive modulation and coding under fading channels. In: 2016 IEEE Wireless Communications and Networking Conference, pp. 1–6 (2016)
Zhao, Y., Chen, Z., Ji, H.: Moore state machine for adaptive modulation and coding in high-speed railway LTE-R systems. Electron. Lett. 52(9), 776–778 (2016)
Zhang, Y., Shen, Y., Zhang, Z., You, X., Zhang, C.: Adaptive spatial modulation combining BCH coding and Huffman coding Tong Xue. In: 2018 IEEE 23rd International Conference on Digital Signal Processing (DSP), pp. 1–5 (2018)
Zeng, R., Liu, T., Yu, X., Zhang, Z.: Novel channel quality indicator prediction scheme for adaptive modulation and coding in high mobility novel channel quality indicator prediction scheme for adaptive modulation and coding in high mobility environments. IEEE Access 7, 11543–11553 (2019)
Azza, M.A., El Yahyaoui, M., El Moussati, A.: Throughput performance of adaptive modulation and coding schemes for WPAN transceiver. In: International Symposium on Advanced Electrical and Communication Technologies (ISAECT), pp. 1–4 (2018)
Muhammad, R., Shikh-Bahaei, M.: Cross-layer combining of truncated ARQ with adaptive modulation and coding in full duplex systems. In: Wireless Advanced (WiAd), pp. 1–6 (2018)
Jiao, W., Ding, H., Wu, H., Yu, G.: Spectrum efficiency of jointing adaptive modulation coding and truncated ARQ with QoS constraints. IEEE Access 6, 46915–46925 (2018)
Shi, F., Yuan, D.: Cross-layer combination of cooperative HARQ with AMC in wireless ad-hoc networks. In: 11th IEEE Singapore International Conference on Communication Systems, pp. 896–900 (2008)
Sunu, C., Aktas, E.: Adaptive modulation and coding (AMC) in cooperative communication channels. In: 22nd Signal Processing and Communications Applications Conference (SIU), pp. 1571–1574 (2014)
Wang, N., Gulliver, A.T.: Packet level analysis for AMC in a wireless cooperative communication system over Nakagami-m fading channels. In: 12th Canadian Workshop on Information Theory, pp. 130–133 (2011)
Harsini, J.S., Farshad, L., Levorato, M., Zorzi, M.: Analysis of non-cooperative and cooperative type II hybrid ARQ protocols with AMC over correlated fading channels. IEEE Trans. Wirel. Commun. 10(3), 877–889 (2011)
Wang, N., Gulliver, T.A.: Cross layer AMC scheduling for a cooperative wireless communication system over Nakagami-m fading channels. IEEE Trans. Wirel. Commun. 11(6), 2330–2341 (2012)
Ramis, J., Femenias, G.: Cross-layer modeling of wireless systems using AMC with cooperative-ARQ error control. In: European Wireless Conference, pp. 1–8 (2012)
Markovic, G.B. Dukic, M.L.: The applicability of cooperative AMC with multiple sensors in dispersive fading channels. In: 2013 21st Telecommunications Forum Telfor (TELFOR), pp. 224–227 (2013)
Bapatla, D., Prakriya, S.: Performance of incremental relaying with an energy-buffer aided relay. In: 2019 IEEE 89th Vehicular Technology Conference (VTC2019-Spring)
Bapatla, D., Prakriya, S.: Performance of energy-buffer aided incremental relaying in cooperative networks. IEEE Trans. Wirel. Commun. 18(7), 3583–3598 (2019)
Liu, D., Zhao, M., Zhou, W.: Energy efficiency optimization in energy harvesting incremental relay system. In: 2018 10th International Conference on Wireless Communications and Signal Processing (WCSP) (2018)
Li, G., Mishra, D., Jiang, H.: Cooperative NOMA with incremental relaying: performance analysis and optimization. IEEE Trans. Veh. Technol. 67(11), 11291–11295 (2018)
Boddapati, H.K., Bhatnagar, M.R., Prakriya, S.: Performance of incremental relaying protocols for cooperative multihop CRNs. IEEE Trans Veh. Technol. 67(7), 6006–6022 (2018)
Proakis, J.: Digital Communications, 5th edn. MacGraw-Hill, London (2007)
Hasna, M.O., Alouini, M.S.: Outage probability of multihop transmission over Nakagami fading channels. IEEE Commun. Lett. 7(5), 216–218 (2003)
Anghel, P.A., Kaveh, M.: Exact symbol error probability of a cooperative network in a Rayleigh fading environnement. IEEE Trans. Wirel. Commun. 3(5), 1416–1421 (2004)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Ben Halima, N., Boujemâa, H. Adaptive cooperation protocol with adaptive modulation and coding. SIViP 15, 323–329 (2021). https://doi.org/10.1007/s11760-020-01757-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11760-020-01757-6