In this paper, we propose a fixed-to-fixed length coding approach to near-lossless compression by leveraging block Markov superposition transmission (BMST) of generalized Reed-Solomon (GRS) codes. To compress non-uniform non-binary sources, we propose two schemes: multi-level coding with a natural mapper and single-level coding with a mapper called sparsifier. The multi-level coding can be proved to achieve the source entropy, while the single-level coding is more suitable for practical use. Both the natural mapper and the sparsifier map non-binary symbols with higher probability to sparser binary vectors of fixed length. When compared with variable-length coding, the most distinguished feature of the proposed coding is that the error propagation caused by a few erroneous bits can be controlled. Even more, the proposed scheme can be easily extended as joint source-channel coding (JSCC) with a wide range of code rates by fixing the input while lengthening the output. Numerical results show that the proposed codes can approach the Shannon limits for transmitting non-uniform sources over noisy channels, providing a universal way to trade off bandwidth and the transmission power.

In this paper, we propose a new coding approach to near-lossless compression of$Q$-ary sources by utilizing a sparsifier and block Markov superposition transmission (BMST) codes. The symbols from a$Q$-ary source are mapped to fixedlength binary vectors by the sparsifier such that the symbols with higher probabilities are mapped to vectors of lower weights. The binary sparse sequences are then compressed in a BMST manner. The most distinguished feature of the proposed source coding is that the error propagation can be mitigated in the presence of noise. Numerical results show that the proposed scheme performs well for$Q$-ary sources, providing a universal but simple way to achieve near-lossless coding at rates approaching the source entropy.

In this paper, we propose to transmit sparse data by exploiting block Markov superposition transmission (BMST) of high-rate low-density parity-check (LDPC) codes. The high-rate LDPC codes are taken as the basic codes to lower down the error floors, while the BMST serves as joint source-channel coding (JSCC). The most distinguished feature of the BMST-LDPC codes is their flexible construction, which requires no complicated optimization and applies to a wide range of code rates. More importantly, embedding LDPC codes into the BMST system allows us to carry the sparsity information by free-ride coding without any loss of code rate, which may find applications especially in the scenario when the source sparsity varies from block to block. Numerical results show that the proposed BMST-LDPC coded system performs well for sparse sources with entropy < 0.5 bits, providing an easy way to trade off between the transmission power and the system bandwidth.

This paper is concerned with the wireless video surveillance systems, which were widely deployed and now augmented by edge computing. Armed with the edge computing, on-device local intelligence algorithms like machine learning (ML) can be utilized to extract specific semantics such as events classification in terms of risk levels which are vital for downstream tasks, say video retrieval. Different from the emergent semantic communications, not only these extracted semantic features (for further use) but also the raw video (by legal requirement) need to be sent to the surveillance center. This application scenario is also different from those for the conventional edge computing. The main objective of this paper is to propose a cost-effective scheme for such extra semantic data transmission in a free-ride way that has mild impact on the existing communication link and requires neither bandwidth expansion nor extra transmission power. The basic idea is to superimpose in the binary field the extra bits on the coded payload data. Numerical results show that, with the 5G low-density parity-check (LDPC) codes, simultaneously transmitting semantic features along with the payload data delivers more reliable semantic features and has a negligible effect on the quality of the payload data.

With industrial control systems (ICSs) commonly connected to the cloud, the security of ICS has received widespread attention. Intrusion detection systems (IDSs) are widely employed to protect ICSs, yet most existing intrusion detection models require expert systems to select the important features and large amount of storage space to store data, both result in increased costs. This paper proposes a preprocessing approach based on hash functions, which does not require prior knowledge and saves a lot of storage space. First, we compute a hash code for each piece of original data with the hash function. Next, the hash codes are converted to decimal and normalised. Finally, a number between 0 and 1 is obtained as a feature for intrusion detection. In experiments, we combine our method with various machine learning algorithms, and extensive experimental results illustrate that our method combined with support vector machine (SVM) and k-nearest neighbor (K-NN) achieve good detection results.

Transformer has recently been introduced into point cloud completion and achieved inspirational performance on 3D point cloud generation. However, the low-level local geometric features are ignored in the existing feature extraction network, and this leads to the loss of geometry in the recovery results. Meanwhile, FoldingNet models that estimate neighborhood information from predicted center points cannot effectively recover geometric information. In this paper, we propose a point cloud completion network based on local geometric feature reuse and neighboring recovery with Taylor approximation (PST-Net). Specifically, a feature extraction network named Skip-DGCNN is constructed to integrate local and global geometric features to reduce the geometry loss during the feature extraction. In addition, we propose a computational model through Taylor approximation to recover the geometry information in the neighborhood of the prediction center. Moreover, we design the TSMB module corresponding to Taylor's approximation to maintain the end-to-end training mode. The proposed method is extensively evaluated and compared with previous methods on three datasets including PCN, ShapeNet-55 and ShapeNet-34. The proposed model outperforms the state-of-the-art (SOTA) and PoinTr on ShapeNet-55 and ShapeNet-34. The complexity analysis on the PCN dataset shows that the number of FLOPs of our approach is 60.79% lower than that of the SOTA. Visual comparisons demonstrate that the proposed method can effectively and accurately complete the geometry of missing parts.

In this paper, we propose a new transmission scheme for the scenario where two nodes attempt to exchange messages and the conventional low-density parity-check (LDPC) codes are implemented for error correction. In the proposed scheme, the ACK/NACK feedback information is transmitted along with the payload data by free-ride codes, while the re-transmitted codewords are superimposed (XORed) on the current codewords, both of which cost neither extra bandwidth nor transmission power. Firstly, we present a syndrome channel model and derive its capacity (referred to as accessible capacity) with a lower bound, implying that the reliable transmission of extra bits (feedback information) is possible. Then, the performance of the extra bits is analyzed by the dependency testing (DT) bound for the syndrome channels. Moreover, motivated by the DT bound, we present a low-complexity DT-like decoder for the free-ride codes. For the superposition retransmission, we present a practical implementation, where those unsuccessfully decoded codewords are sparsely interleaved and superimposed onto the current codewords. In addition, the presented transmission scheme is combined with the conventional hybrid automatic repeat request (HARQ) protocol, resulting in a throughput-enhanced conjunction HARQ scheme. Numerical results show that the word error rate (WER) of the LDPC codes can be significantly reduced by using the presented transmission scheme, but without any extra bandwidth or transmission power. They also show that the presented conjunction HARQ schemes can achieve a throughput improvement up to 80% over fading channels in comparison with the original 5G HARQ scheme.