Journal Article
Machine Learning

Joint Optimization of Radio and Computational Resources for Multicell Mobile-Edge Computing

Stefania Sardellitti(Sapienza University of Rome), Gesualdo Scutari(Purdue University West Lafayette), Sergio Barbarossa(Sapienza University of Rome)
June 1, 2015IEEE Transactions on Signal and Information Processing over Networks919 citations
arXivPDFDOI

919

Citations

46

Influential Citations

IEEE Transactions on Signal and Information Processing over Networks

Venue

2015

Year

Abstract

Migrating computational intensive tasks from mobile devices to more resourceful cloud servers is a promising technique to increase the computational capacity of mobile devices while saving their battery energy. In this paper, we consider an MIMO multicell system where multiple mobile users (MUs) ask for computation offloading to a common cloud server. We formulate the offloading problem as the joint optimization of the radio resources-the transmit precoding matrices of the MUs-and the computational resources-the CPU cycles/second assigned by the cloud to each MU-in order to minimize the overall users' energy consumption, while meeting latency constraints. The resulting optimization problem is nonconvex (in the objective function and constraints). Nevertheless, in the single-user case, we are able to compute the global optimal solution in closed form. In the more challenging multiuser scenario, we propose an iterative algorithm, based on a novel successive convex approximation technique, converging to a local optimal solution of the original nonconvex problem. We then show that the proposed algorithmic framework naturally leads to a distributed and parallel implementation across the radio access points, requiring only a limited coordination/signaling with the cloud. Numerical results show that the proposed schemes outperform disjoint optimization algorithms.

Analysis

Why This Paper Matters

Mobile-edge computing (MEC) is a key enabler for latency-sensitive and computationally intensive applications on mobile devices. This paper addresses a critical gap: most prior work optimized radio or computational resources separately, leading to suboptimal energy efficiency. By jointly designing transmit precoding matrices and CPU-cycle allocation, the authors show that significant energy savings are achievable while meeting strict latency constraints. The work is particularly relevant for dense multicell MIMO systems, which are central to 5G and beyond.

The paper's significance is underscored by its high citation count (919), reflecting its influence on subsequent MEC research. It provides both theoretical insights (closed-form solution for single-user) and practical algorithms (distributed, parallelizable) that balance optimality and scalability.

Technical Contributions

  • Problem formulation: A nonconvex optimization that jointly minimizes user energy consumption over radio precoding matrices and cloud CPU frequencies, subject to per-user latency constraints.
  • Single-user closed-form: Derives the global optimal solution analytically, revealing the trade-off between transmission power and computation speed.
  • Successive convex approximation (SCA): For the multiuser case, the authors develop an iterative algorithm that approximates the nonconvex problem with a sequence of convex subproblems, guaranteeing convergence to a local optimum.
  • Distributed implementation: The SCA framework naturally decomposes across radio access points, requiring only limited signaling with the cloud, making it suitable for large-scale deployments.
  • Numerical validation: Simulations compare the joint approach against disjoint optimization (e.g., separate radio and computation allocation), showing consistent energy reductions.

Results

The paper reports that the proposed joint optimization scheme outperforms disjoint optimization algorithms in terms of total energy consumption. While specific numerical values are not detailed in the abstract, the results demonstrate that the joint approach achieves lower energy for the same latency constraints, especially under high user density or tight delay requirements. The distributed algorithm maintains near-optimal performance with reduced coordination overhead.

Significance

This paper has had a lasting impact on the MEC and wireless resource allocation communities. It established a principled framework for joint communication-computation optimization that has been extended to scenarios with multiple clouds, heterogeneous tasks, and dynamic channels. The distributed SCA approach inspired later work on decentralized optimization in wireless networks. Practitioners can leverage the insights to design energy-efficient offloading strategies for IoT, autonomous systems, and augmented reality applications.

mobile-edge-computingresource-allocationmimoenergy-efficiencyconvex-optimizationdistributed-algorithmslatency-constraints

Related Papers