The development of CubeSat is currently experiencing a rapid increase. One of the growing missions is the space-based Internet of Things (IoT) network using LoRa protocol. The antenna installed on the spacecraft is essential for receiving data from sensors. Compact and adaptable microstrip antennas are commonly used in this design. This study has investigated a microstrip antenna design that incorporated the truncated corner and defected ground structure (DGS) method to improve the antenna performance. Its shape was designed to fit the CubeSat form factor. The antenna is working at 922 MHz of frequency. The experiment result showed that the antenna has return loss of 12.746 dB, 12 MHz of bandwidth, and 1.8 of VSWR. It produces bi-directional beam pattern that has 1.14 dBi of gain with elliptical polarization. The antenna has been integrated with a LoRa module and a range test has been conducted to validate the antenna performance. The result was the antenna was able to communicate with the other LoRa module.
This paper proposes overloaded multiuser multiple-input-multiple-output (MU-MIMO) that is able to increase network throughput by increasing the number of the signal streams to more than that of receive antennas. The spatial filter is employed at each user for such spatial multiplexing. In addition, the proposed overloaded MU-MIMO applies the non-linear precoding with the lattice reduction for improving the transmission performance with low computational complexity. The use of the non-linear precoding can simplify the receiver configuration where amplifiers and the spatial filters are only needed. Computer simulation shows that the proposed overloaded MU-MIMO can double not only the user throughput but also the network throughput compared with the conventional MU-MIMO based on the block diagonalization (BD). In spite of the superior throughput enhancement, the proposed overloaded MU-MIMO achieves 13 dB better transmission performance than the conventional MUMIMO at the BER of $10^{-4}$.
Semantic communications rely on deep neural networks (DNNs) to reduce the amount of transmitted data by only transmitting the semantics of data rather than the whole data, showing the potential on image transmission even in low signal-to-noise-ratio (SNR) conditions. However, the performance deficiency happens once the real-time data do not follow the independent identical distribution (i.i.d) with the training dataset, which results from the poor generalization of DNNs. To tackle this problem, a promising solution is to align the real-time data to follow the similar distribution with the training dataset at the feature level. Thus, we propose a one-shot data adaptive semantic communication (ODASC), where domain adaptation is incorporated as a pre-processing module to cope with domain shift by aligning the distribution between the real-time data and the training dataset. Image transmission is considered as a case study to demonstrate the big plus of ODASC on data recovery and task accuracy.
The utilisation of median filtering (MF), a nonlinear signal processing technique, offers distinct advantages within picture anti-forensics. Consequently, there has been an increased focus on the forensic investigation of MF. However, due to lossy compression, identifying MF in the compressed domain is challenging. Towards this, research presents a novel approach for forensic analysis of MF in compressed images based on utilising deep noise residuals. In this framework, median filtering residuals (MFR) are employed to preprocess the images by passing through two streams. After that, the MFR output is extended to encompass two parallel blocks with different dilation rates to form a fusion feature vector. Further, the MFeRNet framework incorporates convolution, specifically developed to enhance information integration from several streams compared to conventional techniques. The proposed method, MFeRNet, aims to effectively integrate the three-level information of an image and comprehensively extract forensic clues in a compressed scenario. In addition, the experimental results demonstrate that the proposed methodology exhibits superior performance and reduced training time compared to the early reported techniques with equivalent convolution depth.
This paper presents PuFLo, a publish/subscribe forwarding strategy, to deliver content to interested nodes in opportunistic networks. Most of the existing schemes typically use temporal or social knowledge when building routing tables. Instead, PuFLo considers node mobility information when making forwarding decisions. Initially, a node stores the ID of each location it visits in a sequence list. When a node contact occurs, PuFLo selects suitable relays based on two forwarding metrics: mobility similarity and centrality. The former evaluates the mobility diversity of two nodes based on dynamic time warping. The latter assesses node activeness based on Shannon entropy, which reflects the variation of previously visited places. Finally, the two metrics are integrated using the entropy weight method to define the PuFLo forwarding utility. Extensive simulations using the ONE simulator show that PuFLo consistently outperforms Epidemic and SimBet routing in terms of dissemination efficiency, latency, and cost.
