Edge Computing-Enabled Cell-Free Massive MIMO Systems

Sudarshan Mukherjee,Jemin Lee

DOI: https://doi.org/10.48550/arXiv.1908.05857

2019-08-16

Abstract:Mobile edge computing (MEC) has been introduced to provide additional computing capabilities at network edges in order to improve performance of latency critical applications. In this paper, we consider the cell-free (CF) massive MIMO framework with implementing MEC functionalities. We consider multiple types of users with different average time requirements for computing/processing the tasks, and consider access points (APs) with MEC servers and a central server (CS) with the cloud computing capability. After deriving successful communication and computing probabilities using stochastic geometry and queueing theory, we present the successful edge computing probability (SECP) for a target computation latency. Through numerical results, we also analyze the impact of the AP coverage and the offloading probability to the CS on the SECP. It is observed that the optimal probability of offloading to the CS in terms of the SECP decreases with the AP coverage. Finally, we numerically characterize the minimum required energy consumption for guaranteeing a desired level of SECP. It is observed that for any desired level of SECP, it is more energy efficient to have larger number of APs as compared to having more number of antennas at each AP with smaller AP density.

Information Theory,Signal Processing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to improve the success probability of edge computing (SECP) and energy efficiency through the cell - free (CF) massive multiple - input multiple - output (Massive MIMO) system in the mobile edge computing (MEC) environment. Specifically, the paper focuses on the changes in the success probability of edge computing (SECP) under different access point (AP) coverage ranges and the probabilities of offloading tasks to the central server (CS), and explores the impact of these factors on the total network energy consumption.

### The main research questions of the paper include:

1. **Communication performance analysis**:
- Analyzed the uplink and downlink transmission performance of the cell - free massive MIMO system, and derived the successful communication probability (SCMP) using the stochastic geometry method.
- Derived the outage probabilities of the uplink and downlink, and analyzed the relationship between these probabilities and the AP coverage radius.

2. **Computation performance analysis**:
- Considered different types of user tasks with different average processing times on the CS or MEC server.
- Derived the successful computation probability (SCP) under the target computation delay using the queuing theory method, and analyzed the impact of various system parameters (such as AP coverage radius, the probability of offloading tasks to the CS, AP density, etc.) on SCP.

3. **Success probability of edge computing (SECP)**:
- Defined SECP, that is, the probability that a task is successful in both communication and computation.
- Analyzed the impact of the AP coverage radius and the probability of offloading tasks to the CS on SECP, and explored the performance differences between the distributed antenna system and the centralized antenna system in terms of SECP.

4. **Energy consumption analysis**:
- Analyzed the minimum average total energy consumption required to ensure the minimum SECP.
- Explored the impact of AP density on the relationship between the minimum SECP and the total energy consumption.

### Main contributions:

- **Communication performance analysis**: Derived the outage probabilities of the uplink and downlink of the cell - free massive MIMO system, and gave the expression of the successful communication probability (SCMP).
- **Computation performance analysis**: Considered multiple types of tasks and derived the successful computation probability (SCP) using the queuing theory method.
- **Success probability of edge computing (SECP)**: Synthesized the communication and computation performance, derived the expression of SECP, and analyzed its influencing factors.
- **Energy consumption analysis**: Characterized the minimum average total energy consumption required to ensure the minimum SECP, and explored the impact of AP density on energy efficiency.

Through these analyses, the paper aims to provide theoretical support and design guidance for edge computing in the cell - free massive MIMO system, especially in terms of improving the computation success rate and energy efficiency.