Timely Offloading in Mobile Edge Cloud Systems

TL;DR

This work analyzes timely offloading in mobile edge cloud systems under an AoI objective with a cost for MEC usage. It models a generate-at-will source and proves the optimal scheduler has a $z$-dependent age-threshold structure, balancing AoI reduction against MEC costs via a parameter $\lambda$. The authors derive exact expressions for service-threshold metrics, develop a relative value iteration approach to compute the optimal policy, and show that a practical service-threshold policy closely approximates the optimum. The results offer actionable insights for designing MEC-aware schedulers that ensure timeliness while controlling MEC usage in real-world, latency-sensitive applications.

Abstract

Future real-time applications like smart cities will use complex Machine Learning (ML) models for a variety of tasks. Timely status information is required for these applications to be reliable. Offloading computation to a mobile edge cloud (MEC) can reduce the completion time of these tasks. However, using the MEC may come at a cost such as related to use of a cloud service or privacy. In this paper, we consider a source that generates time-stamped status updates for delivery to a monitor after processing by the mobile device or MEC. We study how a scheduler must forward these updates to achieve timely updates at the monitor but also limit MEC usage. We measure timeliness at the monitor using the age of information (AoI) metric. We formulate this problem as an infinite horizon Markov decision process (MDP) with an average cost criterion. We prove that an optimal scheduling policy has an age-threshold structure that depends on how long an update has been in service.

Figures (5)

• Figure 1: System model for offloading: Source generates status updates to a monitor after processing by the local server or MEC.
• Figure 2: Sample variation of the age process at the monitor with $z^*=8$.
• Figure 3: Markov chain for the age threshold policy with $a^*=3$.
• Figure 4: Markov chain of the optimal policy for $\mu=0.5$ and $\lambda=3$. The green lines indicate local service completion. The blue lines indicate that the update is still in service in the local server. The red circles indicate MEC service. $\bar{a}_1=4$, $\bar{a}_2=3.$
• Figure 5: Age vs. frequency of using the MEC for $\mu=0.01$. $a^*$ varies from 1 to 15 for the age threshold policy, and $z^*$ varies from 0 to 9 for the service threshold policy. For the optimal policy, $\lambda\in (0,50)$.

Theorems & Definitions (14)

• Definition 1
• Theorem 1
• proof
• Lemma 1
• Lemma 2
• proof
• Lemma 3: Monotonicity of Value function
• proof
• Lemma 4: Monotonicity of Value function
• proof