Uni-Cycle Genetic Algorithm as an Adaptation Engine for Wireless Channel Equalizers

A recently developed adaptive channel equalizer driven by a so-called Uni-Cycle Genetic Algorithm (UCGA) is examined in the paper. The authors consider different initialization strategies of the iterative process and compare UCGA against the reference Recursive Least Squares (RLS) algorithm in terms of Bit Error Rate (BER) vs. Signal to Noise Ratio (SNR) performance and convergence rate of an adaptive channel equalizer. The results display a reasonable performance gain of UCGA over RLS for most of wireless channel models studied in the paper. Additionally, UCGA is capable of boosting the equalizer convergence. Thus, it can be considered a promising candidate for the future adaptive wireless channel equalizer.

Hybrid Chirp Signal Design for Improved Long-Range (LoRa) Communications

Long-range (LoRa) communication has attracted much attention recently due to its utility for many Internet of Things applications. However, one of the key problems of LoRa technology is that it is vulnerable to noise/interference due to the use of only up-chirp signals during modulation. In this paper, to solve this problem, unlike the conventional LoRa modulation scheme, we propose a modulation scheme for LoRa communication based on joint up- and down-chirps. A fast Fourier transform (FFT)-based demodulation scheme is devised to detect modulated symbols. To further improve the demodulation performance, a hybrid demodulation scheme, comprised of FFT- and correlation-based demodulation, is also proposed. The performance of the proposed scheme is evaluated through extensive simulation results. Compared to the conventional LoRa modulation scheme, we show that the proposed scheme exhibits over 3 dB performance gain at a bit error rate of 10−4.

A Coordinated Direct AF DF Relay-Aided NOMA Framework for Low Outage

This paper investigates the performance of a framework for low-outage downlink non-orthogonal multiple access (NOMA) signalling using a coordinated direct and relay transmission (CDRT) scheme with direct links to both the near-user (NU) and the far-user (FU). Both amplify-and-forward (AF) and decode-and-forward (DF) relaying are considered. In this framework, NU and FU combine the signals from BS and R to attain good outage performance and harness a diversity of two without any need for feedback. For the NU, this serves as an incentive to participate in NOMA signalling. For both NU and FU, expressions for outage probability and throughput are derived in closed form. High-SNR approximations to the outage probability are also presented. We demonstrate that the choice of power allocation coefficient and target rate is crucial to maximize the NU performance while ensuring a desired FU performance. We demonstrate performance gain of the proposed scheme over selective decode-and-forward (SDF) CDRT-NOMA in terms of three metrics: outage probability, sum throughput and energy efficiency. Further, we demonstrate that by choosing the target rate intelligently, the proposed CDRT NOMA scheme ensures higher energy efficiency (EE) in comparison to its orthogonal multiple access counterpart. Monte Carlo simulations validate the derived expressions.

A BERT-Based Two-Stage Model for Chinese Chengyu Recommendation

In Chinese, Chengyu are fixed phrases consisting of four characters. As a type of idioms, their meanings usually cannot be derived from their component characters. In this article, we study the task of recommending a Chengyu given a textual context. Observing some of the limitations with existing work, we propose a two-stage model, where during the first stage we re-train a Chinese BERT model by masking out Chengyu from a large Chinese corpus with a wide coverage of Chengyu. During the second stage, we fine-tune the re-trained, Chengyu-oriented BERT on a specific Chengyu recommendation dataset. We evaluate this method on ChID and CCT datasets and find that it can achieve the state of the art on both datasets. Ablation studies show that both stages of training are critical for the performance gain.

AethelmE, HTML5 Game Engine with Multiple Canvas Elements

Game engine is software which ease the game development. As the processor power technology evolved and the HTML5 (HyperText Markup Language 5) specification are developed, browsers nowadays can natively (without any need for external plug-in) display animations and multimedia files (audio and video) using JavaScript as the programming language. Some of the features which are used in this research are HTML5‘s canvas and audio elements. The problem is that none of the existing free HTML5 game engines is able to support multiple canvas elements. This research will create a game engine, called AethelmE, which support multiple canvas elements as its unique feature. This game engine is also able to support sprite transformation, browsers compatibility, external asset loading, and audio format compatibility. This research successfully resulted in creating an HTML5 game engine which supports multiple canvas elements. It also supports all the scopes, with a small exception on sound format compatibility. Moreover, this research conducted a performance comparison testing of multiple HTML5 game engines, from which can be concluded that multiple canvas elements does not give significant performance gain compared to a single canvas.

Hybrid Chirp Signal Design for Improved Long-Range (LoRa) Communications

Long-range (LoRa) communication has attracted much attention recently due to its application for many Internet of Things applications. However, one of the key problems of the LoRa technology is it is vulnerable to noise/interference due to the use of only up-chirp signals during modulation. In this paper, to solve this problem, unlike the conventional LoRa modulation scheme, we propose a modulation scheme for LoRa communication based on joint up- and down-chirps. A fast Fourier transform (FFT) based demodulation scheme is devised to detect modulated symbols. To further improve demodulation performance, a hybrid demodulation scheme, comprised of FFT and correlation-based demodulation is also proposed. The performance of the proposed scheme is evaluated through extensive simulation results. Compared to the conventional LoRa modulation scheme, we show that the proposed scheme exhibits over 3 dB performance gain at bit error rate of 10^-4.

Hybrid Chirp Signal Design for Improved Long-Range (LoRa) Communications

Long-range (LoRa) communication has attracted much attention recently due to its application for many Internet of Things applications. However, one of the key problems of the LoRa technology is it is vulnerable to noise/interference due to the use of only up-chirp signals during modulation. In this paper, to solve this problem, unlike the conventional LoRa modulation scheme, we propose a modulation scheme for LoRa communication based on joint up- and down-chirps. A fast Fourier transform (FFT) based demodulation scheme is devised to detect modulated symbols. To further improve demodulation performance, a hybrid demodulation scheme, comprised of FFT and correlation-based demodulation is also proposed. The performance of the proposed scheme is evaluated through extensive simulation results. Compared to the conventional LoRa modulation scheme, we show that the proposed scheme exhibits over 3 dB performance gain at bit error rate of 10^-4.

Interference Rejection Combining Approach in Vehicle Communication Systems for Throughput Enhancement

In this paper, we present interference rejection combining scheme for interference suppression in wireless access in vehicular environments (WAVE) system. WAVE system performances depend on interference traffic since various signals and noises are present due to various vehicles on the road. The IRC scheme can minimize the interference presence from the received signal within the massive interference condition, resulting in the substantial gain of signal-to-interference and noise ratios (SINR) and performance. Based on the experiment of our proposed scheme, given the vehicle speed, SINR and different channel condition, our proposed scheme for interference suppression achieved significant improvements by 2 dB SINR performance gain in the low speed condition and above 0.5 dB performance gain at the high speed case. To extend our scheme for the comprehensive analysis, we also produced the vehicle speed and SINR performance map, which showed the performance pattern over vehicle speed and SINR of our scheme.

AdaVQA: Overcoming Language Priors with Adapted Margin Cosine Loss

A number of studies point out that current Visual Question Answering (VQA) models are severely affected by the language prior problem, which refers to blindly making predictions based on the language shortcut. Some efforts have been devoted to overcoming this issue with delicate models. However, there is no research to address it from the view of the answer feature space learning, despite the fact that existing VQA methods all cast VQA as a classification task. Inspired by this, in this work, we attempt to tackle the language prior problem from the viewpoint of the feature space learning. An adapted margin cosine loss is designed to discriminate the frequent and the sparse answer feature space under each question type properly. In this way, the limited patterns within the language modality can be largely reduced to eliminate the language priors. We apply this loss function to several baseline models and evaluate its effectiveness on two VQA-CP benchmarks. Experimental results demonstrate that our proposed adapted margin cosine loss can enhance the baseline models with an absolute performance gain of 15\% on average, strongly verifying the potential of tackling the language prior problem in VQA from the angle of the answer feature space learning.

Non-decreasing Quantile Function Network with Efficient Exploration for Distributional Reinforcement Learning

Although distributional reinforcement learning (DRL) has been widely examined in the past few years, there are two open questions people are still trying to address. One is how to ensure the validity of the learned quantile function, the other is how to efficiently utilize the distribution information. This paper attempts to provide some new perspectives to encourage the future in-depth studies in these two fields. We first propose a non-decreasing quantile function network (NDQFN) to guarantee the monotonicity of the obtained quantile estimates and then design a general exploration framework called distributional prediction error (DPE) for DRL which utilizes the entire distribution of the quantile function. In this paper, we not only discuss the theoretical necessity of our method but also show the performance gain it achieves in practice by comparing with some competitors on Atari 2600 Games especially in some hard-explored games.