Power-Efficient System Design for Cellular-Based Machine-to-Machine Communications

Machine-to-machine (M2M) communications is emerging to be an important service provided over cellular networks by mobile network operators. This is driving the need to optimize the network to efficiently handle growing M2M traffic, which has distinct characteristics compared to human generated traffic. While human consumer traffic requires a small number of long lived sessions, M2M traffic is characterized by a large number of short sessions. Furthermore, M2M constitutes a number of battery powered devices that are static in locations such as basements and underground tunnels and can only communicate with radio signals that are substantially stronger compared to what is required for traditional service. Therefore, recognizing the device transmit power as a critical issue, we focus on the fundamental aspects of system design with the goal of minimizing the total transmit power of all the devices. Specifically, for a variety of coordinated and uncoordinated transmission strategies, we derive results for the optimal transmit power, energy per bit, and the maximum load that can be supported by the base station. Using these results, we show that while the coordinated frequency division multiple access (FDMA) is the best practical strategy overall, uncoordinated code division multiple access (CDMA) is almost as good when the base station is lightly loaded. For coordinated strategies, we derive theoretical guarantees on the performance of simpler strategies, such as equal resource allocation, compared to the optimal performance and prove that the value of optimization is in general not significant in the regime of interest for M2M.