Exploring Distance Query Processing in Edge Computing Environments

Xiubo Zhang,Yujie He,Ye Li,Yan Li,Zijie Zhou,Dongyao Wei,Ryan

2024-03-17

Abstract:In the context of changing travel behaviors and the expanding user base of Geographic Information System (GIS) services, conventional centralized architectures responsible for handling shortest distance queries are facing increasing challenges, such as heightened load pressure and longer response times. To mitigate these concerns, this study is the first to develop an edge computing framework specially tailored for processing distance queries. In conjunction with this innovative system, we have developed a straightforward, yet effective, labeling technique termed Border Labeling. Furthermore, we have devised and implemented a range of query strategies intended to capitalize on the capabilities of the edge computing infrastructure. Our experiments demonstrate that our solution surpasses other methods in terms of both indexing time and query speed across various road network datasets. The empirical evidence from our experiments supports the claim that our edge computing architecture significantly reduces the latency encountered by end-users, thus markedly decreasing their waiting times.

Databases

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

This paper aims to address the issue of accurately processing shortest distance queries in dynamic road network environments. Specifically:

1. **Challenges Faced by Centralized Architectures**:
- With the growth in the number of Geographic Information System (GIS) service users and changes in travel behavior, traditional centralized architectures face increased load pressure and longer response times when handling shortest distance queries.

2. **Proposed New Method**:
- To alleviate these issues, the researchers developed an edge computing framework specifically for handling distance queries and proposed a simple yet effective labeling technique—Border Labeling.
- Additionally, they designed and implemented various query strategies that leverage the capabilities of edge computing infrastructure.

3. **Experimental Results**:
- Experiments show that the proposed solution outperforms other methods in terms of indexing time and query speed across multiple road network datasets.
- Empirical evidence supports that this edge computing architecture can significantly reduce end-user waiting time and improve response speed.

In summary, by introducing an edge computing framework and Border Labeling technique, this paper aims to enhance the efficiency and response speed of distance queries in large-scale dynamic road networks.