Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ReStorEdge: An edge computing system with reuse semantics

Adrian-Cristian Nicolaescu, Spyridon Mastorakis, Md Washik Al Azad, David Griffin, Miguel Rio

TL;DR

ReStorEdge tackles the problem of handling increasingly load-heavy edge computing by exploiting similarity among queries to reuse previous results. It introduces Edge Data Repositories (EDRs) and a central orchestrator that uses Locality Sensitive Hashing to route queries to the most reusable data, reducing edge processing and improving throughput. Four orchestration strategies balance reuse benefits against resource use, and evaluations on real datasets and a large-scale simulator show throughput gains in the 25–33% range, with different strategies excelling across use cases. The work demonstrates a practical path to higher edge QoS through session-less computation reuse and distributed caching, while outlining deployment considerations and directions for energy-aware and cross-provider implementations.

Abstract

This paper investigates an edge computing system where requests are processed by a set of replicated edge servers. We investigate a class of applications where similar queries produce identical results. To reduce processing overhead on the edge servers we store the results of previous computations and return them when new queries are sufficiently similar to earlier ones that produced the results, avoiding the necessity of processing every new query. We implement a similarity-based data classification system, which we evaluate based on real-world datasets of images and voice queries. We evaluate a range of orchestration strategies to distribute queries and cached results between edge nodes and show that the throughput of queries over a system of distributed edge nodes can be increased by 25-33%, increasing its capacity for higher workloads.

ReStorEdge: An edge computing system with reuse semantics

TL;DR

ReStorEdge tackles the problem of handling increasingly load-heavy edge computing by exploiting similarity among queries to reuse previous results. It introduces Edge Data Repositories (EDRs) and a central orchestrator that uses Locality Sensitive Hashing to route queries to the most reusable data, reducing edge processing and improving throughput. Four orchestration strategies balance reuse benefits against resource use, and evaluations on real datasets and a large-scale simulator show throughput gains in the 25–33% range, with different strategies excelling across use cases. The work demonstrates a practical path to higher edge QoS through session-less computation reuse and distributed caching, while outlining deployment considerations and directions for energy-aware and cross-provider implementations.

Abstract

This paper investigates an edge computing system where requests are processed by a set of replicated edge servers. We investigate a class of applications where similar queries produce identical results. To reduce processing overhead on the edge servers we store the results of previous computations and return them when new queries are sufficiently similar to earlier ones that produced the results, avoiding the necessity of processing every new query. We implement a similarity-based data classification system, which we evaluate based on real-world datasets of images and voice queries. We evaluate a range of orchestration strategies to distribute queries and cached results between edge nodes and show that the throughput of queries over a system of distributed edge nodes can be increased by 25-33%, increasing its capacity for higher workloads.
Paper Structure (20 sections, 10 figures, 3 tables, 4 algorithms)

This paper contains 20 sections, 10 figures, 3 tables, 4 algorithms.

Table of Contents

  1. Introduction
  2. Prior Related Work and Use Cases
  3. Edge-based Data Storage and Computation Offloading
  4. Computation Deduplication and Reuse
  5. Locality Sensitive Hashing and Network Services
  6. Use Cases
  7. System Design
  8. Definitions of terms
  9. Description of ReStorEdge Operational Workflow
  10. Query Routing
  11. Storage Similarity and Reuse
  12. Orchestration Algorithms
  13. Evaluation
  14. Setup and Assumptions
  15. EDR-based ReStorEdge Prototype Setup and Metrics
  16. ...and 5 more sections

Figures (10)

  • Figure 1: Overview of ReStorEdge Workflow
  • Figure 2: Processing vs. Reuse timings for different datasets.
  • Figure 3: Throughputs for different rates with MNIST Dataset.
  • Figure 4: Throughputs for different rates with Traffic Detection Dataset.
  • Figure 5: Throughputs for different rates with Alexa Dataset.
  • ...and 5 more figures