Performance analysis of cache-aided UAV relaying networks

Abstract

In this work, we analyze the effect of relaying network where unmanned aerial vehicles (UAVs) act as cache-aided relay nodes. Unmanned aerial vehicle (UAV) has high mobility, which can be flexibly scheduled when unable to communicate so that it can act as a wireless relay to assist communication. It can provide a more flexible deployment mode in different application scenarios to improve the performance of communication. What is more, how to transform the dynamic characteristics of UAVs into channel characteristics is also a challenge. To be more specific, UAVs act as decoded-and-forward (DF) relays to assist the wireless communication between the source and destination, and the speed and flying height remains unchanged. The UAVs are equipped with a wireless cache, which can provide reliable services and expand the channel capacity to reduce the outage probability of the network. Experimental and simulation results verify the effect of UAVs with cache on the outage performance in relaying networks.

Introduction

Since wireless relay communication was proposed, the wireless relay model has been applied in various scenarios to improve the channel capacity [1], [2], [3]. In the development process of wireless relay networks, there are many types of researches on it from different angles. Cooperative diversity using relay signals can resist the fading caused by multipath propagation in wireless networks [4], [5], [6]. In these works, researchers have studied the cooperative diversity realized by relay signals, which can resist the fading caused by multipath propagation in wireless networks. Besides, some works has been made on the effect of energy harvesting on relaying networks. UAV can be used as a relay node to connect remote nodes and gain significant throughput gain, which is attributed to its mobility, flexible deployment, large enough cache capacity, and power consumption. In [7], the author provides a detailed survey on the measurement methods for UAV channel modeling using a low altitude platform and points out that accurate channel characteristics are the key to the optimization and design of UAV communication performance. Then the influence of ground user mobility, propagation environment, and channel fading on UAV communication outage performance is analyzed in [8]. In [9], the author analyzed the effects of outdated antenna selection on cache-aided unmanned aerial vehicle (UAV) relay networks.

With the great development of caching and computing resources, most vehicles, base stations, and terminals have the corresponding capabilities. In [10], the authors take advantage of the popularity of cache in infrastructure, to improve the performance of the communication networks. What is more, Vehicle-to-vehicle (V2V) and Vehicle-to-infrastructure (V2I) are proposed to make the links between terminal to terminal more diversified and make full use of idle resources to improve the quality of service of wireless communication [11], [12]. Due to the limited communication capacity of infrastructure and the mobility and selfishness of vehicles, how to motivate vehicles to participate in the data dissemination process has become a challenge, the author in [13] proposed a new content distribution framework based on edge computing to solve this problem. In [14], the author studies a relaying system, by solving a non-convex problem with information causality constraints, the location, node power, and bandwidth allocation of UAV relay are optimized. Line of sight channels make UAV transmission easy to eavesdrop, authors in [15] propose a novel scheme to guarantee the security of UAV-relayed wireless networks with caching via jointly optimizing the UAV trajectory and time schedule. What is more, energy efficiency has become a significant metric for UAVs owning to their limited energy. Thus, authors in [16] aim to maximize the energy efficiency for mmWave-enabled NOMA-UAV networks by optimizing the UAV placement, hybrid precoding and power allocation. In addition, there are many works in different application scenarios, using the characteristics of UAV to design the corresponding ad hoc network or scheme to solve the corresponding problems [17].

Although the mobility and flexible deployment of UAV can make the wireless communication network obtain significant throughput gain, the interference caused by a large number of wireless spectrum reuse and noise will still have adverse effects on wireless communication. The equipment of cache is considered to be an effective way to release the potential of wireless communication network, which can provide services with low latency and high reliability. Yang et al. [18] proposed a heterogeneous network based on cache content delivery, where base stations (BSs), relays, and device-to-device (D2D) pairs are included. The pre-stored content, which is sent out from the adjacent point, can be either relay-to-device (R2D) or D2D. In addition, researchers have further studied a UAV-aided decoding-and-forwarding (DF) relay communication system for downlink maritime communication, by taking advantage of cache’s low latency, high reliability, and the flexibility of UAV deployment. Thus it can be seen that cache has rich application scenarios in wireless communication, and it will also play a very important role in beyond 5G (B5G) communications.

In this paper, we investigate the effect of cache-aided UAV on outage performance in relaying networks, where multiple UAVs equipped with wireless cache act as DF relays. In our system model, it is assumed that UAVs move from source to destination at a constant speed and flight altitude, and assist data transmission from source to destination. In order to illustrate the effect of cache-aided UAV on the performance of wireless communication system, we design several optimization criteria for the UAV network equipped with cache and without cache. Specifically, According to different optimization objectives, we design the corresponding optimization criteria for two UAV relay networks and derive the analytical and the asymptotic expressions of outage probability. Finally, we give the simulation results to verify the correctness of the research results and the advantages of cache-aided UAV in the wireless communication system. The main contributions of the paper are summarized as follows.

$•$ In this paper, we consider a two-hop relaying network with $N$ UAVs, where UAVs act as relay nodes. Noise and interference can affect communication at both relays and the destination. Therefore, in this paper, we design corresponding optimization criteria for first-hop and second-hop respectively. Through the experimental results, we can see the effect of UAVs and using different optimization criteria in different positions on the system performance.

$•$ To intuitively reflect the advantages and disadvantages of our system, we derive the analytical expression of outage probability of relaying network using various optimization criteria. Besides, in order to get some insights of the system, the asymptotic expression of outage probability is also derived.

$•$ From the preliminary experimental results, we can determine the location of UAV, where the optimization effect of each criterion can reach the best, and determine the position in the follow-up experiments, which is further illustrated by simulation. Moreover, the curve fitting of all the simulation results with the mathematical expression also shows the correctness of our work.

The remainder of the paper is organized as follows. In the following section, we provide the relaying network model and discuss it in detail. The UAV relay selection criteria are proposed, and the relevant symbols are illustrated in Section 3. In Section 4, the analytical and the asymptotic expressions of the outage probabilities of each criterion are derived. Finally, Section 5 gives the simulation results, and the improvement of system performance by cache-aided UAV is reflected, and conclusions are provided in Section 6.

Notations

Let $X\sim \mathcal{CN}(0,{\alpha }^2)$ be a circularly symmetric complex Gaussian random variable (RV) $X$ with zero mean and variance ${\alpha }^2$. Let $f_X\left(x\right)$ denote the probability density function (PDF) of the RV $X$. And the operation $Pr\left(\cdot \right)$ returns probability.