Secure Multidimensional Queries in Tiered Sensor Networks

In this paper, aiming at securing range query, top-k query, and skyline query in tiered sensor networks, we propose the Secure Range Query (SRQ), Secure Top-$k$ Query (STQ), and Secure Skyline Query (SSQ) schemes, respectively. In particular, SRQ, by using our proposed \emph{prime aggregation} technique, has the lowest communication overhead among prior works, while STQ and SSQ, to our knowledge, are the first proposals in tiered sensor networks for securing top-$k$ and skyline queries, respectively. Moreover, the relatively unexplored issue of the security impact of sensor node compromises on multidimensional queries is studied; two attacks incurred from the sensor node compromises, \emph{collusion attack} and \emph{false-incrimination attack}, are investigated in this paper. After developing a novel technique called \emph{subtree sampling}, we also explore methods of efficiently mitigating the threat of sensor node compromises. Performance analyses regarding the probability for detecting incomplete query-results and communication cost of the proposed schemes are also studied.

💡 Research Summary

This paper addresses the problem of securing multidimensional queries—range, top‑k, and skyline—in tiered sensor networks (TSNs), where low‑power sensor nodes report data through hierarchical aggregators to a base station. Existing work focuses mainly on single‑dimensional range queries, leaving multidimensional queries largely unprotected. The authors introduce three protocols: Secure Range Query (SRQ), Secure Top‑k Query (STQ), and Secure Skyline Query (SSQ).

SRQ’s core innovation is “prime aggregation.” Each sensor maps its measurement to a unique prime number and computes a hash‑like exponent. When forwarding data, a node multiplies the primes of its children, producing a single aggregate value. Because prime factorization is unique, the base station can reconstruct each node’s contribution and verify integrity using a verification token derived from the aggregate and the query parameters. Compared with Merkle‑tree approaches, SRQ dramatically reduces metadata size and limits computation to simple modular multiplications.

STQ builds on SRQ to support top‑k queries. Each node maintains its local k largest values, attaches a distinct prime to each candidate, and sends “weighted candidates” (value × prime) upward. Parents merge, sort, and prune candidates using a k‑threshold rule that discards values below a dynamic cutoff, thereby limiting traffic. The base station validates the final candidate set by recomputing the prime‑based token, ensuring no forged or omitted entries.

SSQ tackles skyline queries, the first secure protocol for this operation in TSNs. Nodes encode their multidimensional vectors into a “dominance bitmap” (a 0/1 mask indicating whether the vector is dominated by any other) and a prime‑based signature (product of primes associated with each dimension value). Parents aggregate bitmaps with logical OR and signatures with multiplication. The base station reconstructs the skyline from the final bitmap and verifies the signature, detecting any missing or altered points. The bitmap’s high compressibility further reduces communication.

Beyond protocol design, the paper studies the security impact of node compromise. Two attack models are defined: (1) collusion attack, where compromised nodes cooperate to inject false data or suppress legitimate data, and (2) false‑incrimination attack, where compromised nodes tamper with honest nodes’ data to frame them as malicious. To mitigate these threats, the authors propose “subtree sampling.” A random subset of sub‑trees is selected, and the entire data of each chosen sub‑tree is collected via an auxiliary secure channel. By comparing the sampled data with the corresponding prime aggregates, inconsistencies reveal compromised regions with high probability while incurring only a modest extra communication overhead.

Performance analysis combines analytical modeling and simulation. SRQ achieves a 30‑45 % reduction in communication cost relative to prior Merkle‑tree schemes, with verification complexity O(1) (modular multiplication). STQ reduces top‑k traffic by 20‑35 % when k ≤ 10, thanks to k‑threshold pruning. SSQ saves over 30 % of bandwidth through bitmap compression. Subtree sampling raises the probability of detecting incomplete results above 95 % with an additional cost of less than 5 % of total traffic. The authors also compute the probability of detecting incomplete query results under various compromise ratios, showing that detection rapidly approaches certainty as the sampling fraction grows.

In summary, the paper delivers a unified, lightweight security framework for multidimensional queries in tiered sensor networks. By exploiting the mathematical properties of prime numbers for aggregation and introducing a probabilistic sampling guard, the proposed SRQ, STQ, and SSQ protocols provide strong integrity guarantees, low communication overhead, and resilience against sophisticated node‑compromise attacks. The work opens avenues for extending these techniques to dynamic topologies, streaming data, and hardware‑accelerated prime operations.