Enabling Skip Graphs to Process K-Dimensional Range Queries in a Mobile Sensor Network

Gregory J. Brault,Christopher James Augeri,Barry E. Mullins,Rusty O. Baldwin,Christopher B. Mayer

DOI: https://doi.org/10.1109/NCA.2007.18

2024-11-26

Abstract:A skip graph is a resilient application-layer routing structure that supports range queries of distributed k-dimensional data. By sorting deterministic keys into groups based on locally computed random membership vectors, nodes in a standard skip graph can optimize range query performance in mobile networks such as unmanned aerial vehicle swarms. We propose a skip graph extension that inverts the key and membership vector roles and bases group membership on deterministic vectors derived from the z-ordering of k-dimensional data and sorting within groups is based on locally computed random keys.

Information Theory,Discrete Mathematics,Data Structures and Algorithms,Networking and Internet Architecture

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to efficiently process k - dimensional range queries in mobile sensor networks (such as drone swarms). Specifically, the author proposes an extended skip graph structure to improve the query performance of distributed multi - dimensional data, especially in time - critical application scenarios.

### Main Problems

1. **Limitations of Existing Methods**
- Standard skip lists group nodes by randomly calculating member vectors and are sorted according to deterministic key values. This may lead to poor query performance in some cases.
- Although multi - dimensional skip lists can handle multi - dimensional data, their construction and query processes still rely on one - dimensional linearized key values, which may affect query efficiency.

2. **Optimizing Query Performance**
- The author hopes to improve the performance of multi - dimensional range queries by improving the skip list structure, especially by inverting the roles of key values and member vectors.
- The goal is to reduce query response time, especially in mobile environments, such as sensor networks in drone swarms.

### Solution

The author proposes a multi - dimensional inverted skip graph, with the following main features:

- **Inversion of Key Values and Member Vectors**: The key values of nodes are randomly generated, while the member vectors are determined based on the multi - dimensional data of nodes (such as geographical coordinates).
- **Z - order Linearization**: Use the Z - order curve to linearize multi - dimensional data into one - dimensional key values for effective sorting and querying in the skip list.
- **Query Mechanism**: Queries start from the lowest level and gradually match longer prefixes upwards until a list of completely matching nodes is found or the highest level is reached.

### Expected Effects

Preliminary results show that under certain query conditions, especially in mobile environments, this new skip list structure can significantly reduce the total number of messages, thereby improving query performance.

### Formula Representation

- **Z - order Linearization Formula**
Let \( (x_1, x_2, \ldots, x_k) \) be k - dimensional coordinates, then its Z - order key value \( z \) can be calculated by bit interleaving:
\[
z=\text{interleave}(x_1, x_2, \ldots, x_k)
\]
For example, for two - dimensional coordinates \( (6, 1) \), their binary representations are \( 110_2 \) and \( 001_2 \) respectively, then the Z - order key value is:
\[
z = 101001_2
\]

In this way, the paper aims to provide a more efficient multi - dimensional range query mechanism, especially suitable for large - scale distributed systems in mobile sensor networks.