IRS-Assisted Wireless Sensor Networks With Hybrid TDMA-NOMA

This paper investigates the energy harvesting (EH) and information decoding (ID) capabilities of wireless sensor networks (WSN) employing a hybrid time division multiple access (TDMA) and non-orthogonal multiple access (NOMA), i.e., the WSN with a hybrid TDMA-NOMA system. The system is assisted by a set of intelligent reflecting surface (IRS) units. In this configuration, the sensors are divided into a set of clusters, with each cluster consisting of two sensors. The available transmission time is equally divided into two phases: the downlink and uplink phases. In the downlink phase, the downlink time slot is dynamically split between the clusters into sub-time slots, where each sub-time slot is further divided into two slots, namely the wireless information transfer (WIT) and the wireless energy transfer (WET) slots. In the WIT slot, the sensors in the clusters use the received signal for ID, while the WET time slot is reserved for EH. However, in the uplink phase, the uplink time slot is also dynamically split into a set of slots, where each time slot is dedicated to assisting the uplink transmission from the sensors in each cluster to the base station (BS). To demonstrate the performance of such a system, we formulate a resource allocation framework that aims to minimize the total transmit power in the system while meeting a set of quality of service (QoS) requirements. Specifically, the total transmit power accounts for the downlink and uplink power, while the QoS requirements include a pre-defined minimum downlink data rate, minimum harvested energy for each sensor in the system, and minimum uplink data rate requirements. However, the joint nature of the optimization parameters in the downlink and uplink phases, namely power allocations in the downlink and uplink, time durations, and phase shift reflecting coefficients of the IRS units, as well as the non-convexity of the problem, introduces additional challenges in solving the formulated power minimization problem. To overcome these challenges, an iterative algorithm is proposed to solve the formulated optimization problem. To demonstrate the potential benefits of the proposed configuration, we present a set of simulations that evaluate its performance against two benchmarks: the IRS-free hybrid TDMA-NOMA system and the IRS-assisted hybrid TDMA-NOMA system with equal time allocations.