Distributed data storage (DDS) can assist compressive crowdsensing (CCS) to solve the challenge of temporary data storage in the network. Block compressive sensing effectively addresses DDS of large spatiotemporal data from crowdsensing, allowing efficient reconstruction at low storage and computational costs. However, when the data is unevenly distributed over the sensing area, existing algorithms ignore the variability of information between blocks and still require the central server to uniformly collect samples from each block stored on the mobile device, leading to a reduction in overall reconstruction accuracy. To address this, we propose an adaptive sampling allocation strategy that deeply analyzes the statistical information of each block which can help the central server to collect the number of measurements for each block adaptively to improve the sampling quality. Additionally, we consider the correlation between blocks and use a global denoising strategy to further improve the reconstruction accuracy. Experimental results demonstrate that, compared to the state-of-the-art DDS-CCS algorithm, our proposed adaptive sampling allocation with a joint-denoising mechanism significantly improves the accuracy of the information-rich blocks that most affect the global accuracy, and hence the global reconstruction accuracy. which also remains robust to different block sizes and exhibits improved stability.

Distributed data storage (DDS) in mobile crowdsensing (MCS) systems has recently gained popularity. Data should be briefly saved on participants’ mobile devices before being gathered once the centralized cloud servers resume normal operations. For MCS systems, the existing DDS strategies briefly considered reconstructing the scene as precisely as possible without thinking about the costs of each step. However, our goal is to obtain a sufficiently accurate approximation of the sensing data from mobile participants with as few costs as possible. We note a crucial observation: when a specified number of participants have been transmitted to a central server, the sensing data has already been well reconstructed, and the accuracy advancement with additional transmitted participants is minimal. In our scheme, two stopping criteria are proposed for DDS in compressive MCS, which aims to enhance recovery performance while reducing the costs of the whole process. In the first stopping criterion, we established a rule to stop the continued recruitment of participants. The algorithm adaptively increases the number of participants until the reconstruction accuracy meets the requirement. Another stopping criterion of the reconstruction algorithm is designed to find a more accurate number of iterations than the original. The experiment results demonstrate that the first stopping criterion can reduce participants’ collection while obtaining an approximate value. The second stopping criterion assists the reconstruction algorithm in terminating at a more appropriate number of iterations, saving computing costs while ensuring accuracy.

We investigate the problem of distributed networked storage with compressive sensing in wireless sensor networks, and a compressive storage scheme for local data query is proposed. Specifically, we propose a simple but efficient one-step data dissemination strategy, and the dissemination cost is reduced dramatically. We further present a lazy-encoding algorithm, using which the local data are capable of being reconstructed without recovering the global data field if not necessary. Thus the decoding ratio decreases significantly. Experiments using real sensor data show that the proposed scheme achieves far better local data recovery performance compared to the existing ones.

Block based compressed sensing (BCS) is a novel framework in image signals sampling and recovery due to its advantages in terms of both low sampling burden and lightweight recovery complexity. In this paper, we propose a novel asymmetric BCS scheme to further improve the image recovery accuracy. In the sampling process, image blocks are partitioned into smaller sub-blocks, and those small sub-blocks are used to allocate sampling resources. In the recovery process, the small sub-blocks with similar feature information are assembled into virtual blocks with larger size, and the corresponding transforming coefficients are then more compressible. The proposed scheme improves the recovered images from the fairer resources allocation and much greater compressibility. The experimental results demonstrate that, compared to the existing BCS approaches, our proposed scheme has higher recovery quality, without increasing sampling and recovery complexity.