Analysis and Optimization of Age of Information for Real-Time Communications

Abstract

In order to meet the increasing demand for wireless communication, the industry and academia in the telecommunications sector have shifted their interests towards exploring the next generation communication technologies, known as the sixth generation (6G) communications. In the envisioned 6G communications, many applications and services rely on real-time communications where timely information delivery is essential. To characterize the freshness of delivered information, an emerging performance metric named age of information (AoI) has been introduced, which is defined as the elapsed time since the last successfully received packet was generated by the transmitter.

Motivated by the importance of AoI, this thesis aims to analyze and optimize the AoI performance in wireless networks. Specifically, this thesis investigates and addresses the following issues: (i) Transmission scheduling policy design in multi-user systems (Chapter 2), (ii) Offloading strategy design in multi-user mobile edge computing (MEC) systems (Chapter 3), (iii) Analysis of AoI performance for different encoding schemes and non-linear AoI performance in short packet communication systems (Chapter 4), and (iv) Optimal transmission selection in short packet communications (Chapter 5). The contributions made in this thesis are summarized as follows:

First, we deign scheduling policies of multi-user systems to optimize the AoI performance in Chapter 2. In this chapter, we design two different scheduling policies, where the first one is a decentralized uplink packet arrival-independent scheduling policy and the second one is a centralized downlink packet arrival-dependent scheduling policy. In particular, we first propose a repeated random access based scheduling policy, where each user equipment (UE) generates its packet and then randomly selects one channel from multiple orthogonal frequency channels to perform repeat transmission of its packet across multiple time slots. We then propose a Whittle index based scheduling policy to minimize the average general AoI penalty in a downlink multi-user system with stochastic packet generation pattern and unreliable transmission links. We formulate the transmission scheduling problem as an average cost constrained Markov decision process problem. Through introducing the service charge, we derive the approximated expression for the Whittle index, based on which we design the scheduling policy.

Second, we investigate the offloading strategy of a multi-user MEC system to achieve the optimal AoI performance in Chapter 3. We analyze the average AoI and the average peak AoI (PAoI) of a multi-user MEC system. In this MEC system, we derive closed-form expressions for the average AoI and average PAoI for three computing schemes. To reduce the complexity of the average AoI expression, we derive upper and lower bounds on the average AoI, which allow us to explicitly examine the dependence of the optimal offloading decision on the MEC system parameters.

Third, we investigate the AoI performance of short packet communication systems in Chapter 4. In this chapter, we first analyze the impact of different encoding schemes on the AoI performance in a short packet communication system. We then analyze the non-linear AoI performance in a short packet communication system.

Finally, we analytically decide whether the broadcast transmission scheme or the unicast transmission scheme achieves the optimal AoI performance of a multi-user system in Chapter 5. In this system, we consider two transmission schemes, i.e., the broadcast transmission scheme and the unicast transmission scheme. For both transmission schemes, we examine three packet management strategies. We derive closed-form expressions for the average AoI achieved by two transmission schemes with three packet management strategies. Based on them, we compare the AoI performance of two transmission schemes in two systems. Show more