Outage Performance of CDF-Based Scheduling in Downlink and Uplink NOMA Systems

The scheduling strategy is a vital precondition for achieving remarkable performance benefits offered by non-orthogonal multiple-access (NOMA) schemes. In this article, the cumulative distribution function (CDF)-based scheduling for NOMA (CS-NOMA) networks with randomly deployed users is investigated. We consider two practical successive interference cancellation (SIC) restrictions, namely imperfect SIC and SIC power difference constraint, in both downlink and uplink transmissions. Exact analytical expressions for the outage probabilities of the two scheduled users are derived in both fixed power allocation (FPA) and cognitive-radio-inspired power allocation (CPA) scenarios. To get more insights, high signal to noise ratio (SNR) approximations or bounds of the outage probabilities are derived, and then be utilized to analyze the achieved diversity orders. Assuming the number of near and far users are respectively represented as $K$ and $B$. Results reveal that, in downlink transmission, the far user can achieve a diversity order of $B$ in both FPA and CPA policies, while the near user's diversity order will be reduced from $K$ (in FPA) to $\text{min}\lbrace K,B\rbrace$ (in CPA). However, in uplink transmission, the achieved diversity orders of the near and far users will be increased from zeros to $K$ and $\text{min}\lbrace K,B\rbrace$, respectively. Simulation results are provided to validate the accuracy of the analytical expressions.

I. Introduction

The spectral efficiency and connectivity density of the next generation systems are expected to significantly increase. To this end, non-orthogonal multiple access (NOMA) has attracted much attention in recent years, which is capable of achieving high spectral efficiency and simultaneously serving multiple users with a single resource block [2]. It has been widely recognized as one of the most promising technologies in the next generation mobile communication systems [3]. In fact, NOMA has been included in the 3 rd generation partnership project (3GPP) 5 G New Radio (NR) systems to increase capacity for eMBB, and also enable grant-free transmission for URLLC and mMTC [4]. In general, NOMA can be divided into two major categories, namely code-domain NOMA (CD-NOMA) and power-domain NOMA (PD-NOMA) [5]. This paper focuses on the investigation of PD-NOMA, and unless otherwise stated, we utilize NOMA to represent PD-NOMA in the rest of the paper. The main principle of NOMA is that the signals of different users are superimposed and transmitted with distinct power levels in the same time/frequency/code resource, and the multi-user detection algorithm, i.e., successive interference cancellation (SIC), is employed at the receivers to extract the desired signals.