ENHANCING ENERGY EFFICIENT TASK CACHING AND OFFLOADING IN MOBILE EDGE COMPUTING

In response to the escalating global energy crisis, rising energy costs, and heightened awareness of the environmental impact of carbon emissions, this work addresses the imperative to reduce the energy consumption associated with MEC (Mobile Edge Computing, a distributed computing paradigm that processes data closer to the network edge, enhancing the efficiency and latency of mobile networks by deploying computational resources at the edge of the network infrastructure). Mobile Edge Computing (MEC) enables the prolongation of battery life in mobile devices and supports the execution of computationally intensive applications at the edge. This can be achieved by offloading these tasks to a server deployed near the base station and/or by directly caching them. Previous work either focuses on one of these two strategies or formulates optimization problems that are hard to solve, proposing suboptimal solutions.

In response to the escalating global energy crisis, rising energy costs, and heightened awareness of the environmental impact of carbon emissions, this work addresses the imperative to reduce the energy consumption associated with MEC (Mobile Edge Computing, a distributed computing paradigm that processes data closer to the network edge, enhancing the efficiency and latency of mobile networks by deploying computational resources at the edge of the network infrastructure). Mobile Edge Computing (MEC) enables the prolongation of battery life in mobile devices and supports the execution of computationally intensive applications at the edge. This can be achieved by offloading these tasks to a server deployed near the base station and/or by directly caching them. Previous work either focuses on one of these two strategies or formulates optimization problems that are hard to solve, proposing suboptimal solutions. The work presented in this thesis introduces a linearized model that transforms a complex mixed-integer nonlinear programming problem into a more tractable linear form. This model enables the derivation of optimal solutions for large-scale instances, making it suitable for real-world applications. Additionally, the thesis proposes heuristic algorithms that offer near-optimal solutions while significantly reducing computational complexity. Through extensive simulations, the thesis demonstrates that the proposed methods not only achieve substantial energy savings but also scale effectively with increasing network sizes.