Systems and Solutions Integration

Link adaptation mechanisms are viewed as one of the key techniques for increasing the diversity order, robustness and effectiveness of a wireless communication system. In this technical report, we present a novel algorithm in which we co-design Adaptive Modulation and Coding (AMC), Hybrid ARQ (HARQ) and power control in order to enhance the energy efficiency of packet delivery, while meeting the QoS constraints. Telecommunication has experienced tremendous success causing proliferation and demand for ubiquitous heterogeneous broadband mobile wireless communications. Up to now, innovation has targeted to improve wireless networks coverage and capacity while meeting the QoS for users admitted in the system. With this report, we aim to investigate how adaptive mechanisms can be exploited to improve the energy efficiency of wireless communications. Specifically, we focus on Adaptive Modulation and Coding (AMC), Hybrid Automatic Repeat Request (HARQ) and power control algorithm, in order to improve the energy efficiency of transmission. We propose to adaptively achieve the optimization tradeoff between power, spectral efficiency and delay, by adapting at each transmission attempt of a packet, the risk of unsuccessful packet delivery. QoS constraints of heterogeneous application services are taken into account in terms of residual packet error rate after a pre-fixed number of maximum retransmission attempts. We call the proposed algorithm Variable PER Adaptation. Fig. 1: Energy Efficiency gain(%) Vs UE Downlink traffic compared to PERtg = 10% for Residual PER, PERres = 0.01 Fig. 1 shows the Energy Efficiency (EE) gain Vs UE downlink traffic compared to fixed Packet Error Rate (PER) target, PERtg = 10%. The residual PER, PERres for the application is 0.01. We plot the results of EE gain Vs User Equipment (UE) downlink traffic for users with varying distance, D from Base-Station (BS). We can see from fig. 1 that the UE closer to the BS gains much more in terms of EE than the UE farther away from BS. The reason for this behavior is that whenever there is a better channel compared to the spectral efficiency requirement of the user based on downlink traffic, we have more degrees of freedom to optimize the target packet error rate. Hence, the UEs which have better momentarily channel are more suited to gain from the variable PER adaptation. The EE gain for low load scenarios especially for UEs closer to the base-station can be as high as 20%. The EE for high load scenarios is very little due to the fact that high data rate requirement from the application limits the degree of freedom available to the algorithm to reduce power consumption. We also plot the gain in average number of retransmissions Vs UE downlink traffic for the two schemes in fig. 2. The variable PER adaptation increases the delay of packets, but, this delay is acceptable as long as its below the tolerance level desired by the application. The variable PER algorithm exploits the classical tradeoff between delay and energy consumption. Fig. 2: Average number of retransmissions gain(%) Vs UE downlink traffic for PERres = 0:05 References: [1] E. Calvanese Strinati and R. Gupta, “Adaptive packet error rate target for energy efficient packet delivery”, EUSIPCO 2011, Barcelona, Spain.