With the rapid development of Internet technology, network attacks have become more frequent and complex, and intrusion detection has also played an increasingly important role in network security. Intrusion detection is real‐time and proactive, and it is an indispensable technology under the diversified trend of network security issues. In terms of network security, neural networks have the characteristics of self‐learning, self‐adaptation, and parallel computing, which are very important in intrusion detection. This paper combines back propagation neural network (BPNN) and elite clone artificial bee colony (ECABC) to propose a new ECABC‐BPNN, which updates and optimizes the settings of traditional BPNN weights and thresholds. Then, apply ECABC‐BPNN to network intrusion detection. Use the attack data samples of KDD CUP 99 and water pipe for attack classification experiments using GA‐BPNN, PSO‐BPNN, and ECABC‐BPNN. The results show that the ECABC‐BPNN proposed in this paper has an accuracy rate of 98.08% on KDD 99 and 99.76% on water pipe data. ECABC‐BPNN effectively improves the accuracy of network intrusion classification and reduces classification errors. In addition, the time complexity of using ECABC‐BPNN to classify network attacks is relatively low. Therefore, ECABC‐BPNN has superior performance in network intrusion detection and classification.
Multi-path AODV routing protocol(MP-AODV) uses the repetitive RREQ packets received by the nodes in AODV routing protocol to build the multi-path routes.Simulation based on NS2 software indicates that both datagram successful transmission rate and delay time is improved,being compared with the AODV routing protocol.
Large-scale wireless sensor networks (LSWSNs) are currently one of the most influential technologies and have been widely used in industry, medical, and environmental monitoring fields. The LSWSNs are composed of many tiny sensor nodes. These nodes are arbitrarily distributed in a certain area for data collection, and they have limited energy consumption, storage capabilities, and communication capabilities. Due to limited sensor resources, traditional network protocols cannot be directly applied to LSWSNs. Therefore, the issue of maximizing the LSWSNs’ lifetime by working with duty cycle design algorithm has been extensively studied in this paper. Encouraged by annealing algorithm, this work provides a new elite adaptive simulated annealing (EASA) algorithm to prolong LSWSNs’ lifetime. We then present a sensor duty cycle models, which can make sure the full coverage of the monitoring targets and prolong the network lifetime as much as possible. Simulation results indicate that the network lifetime of EASA algorithm is 21.95% longer than that of genetic algorithm (GA) and 28.33% longer than that of particle swarm algorithm (PSO).
The wormhole attack is a severe attack in Wireless Mesh Networks (WMNs). It involves two or more wormhole endpoints colluding to capture traffic from one place in the network and replay it to another faraway place through a secret tunnel, so as to distort network routing. It may lead to even more serious threats such as packet dropping and denial of service (DoS). Although a lot of works have been done on detecting wormhole attacks, few of them actually evaluated their solutions on a testbed to consider the real network conditions. In this paper, we set up a WMN testbed for studying wormhole attacks to fill this gap. Some existing approaches used RTT to detect wormhole attacks. However, from both theoretical analysis and experimental results, we observed that the standard deviation of round trip time (stdev(RTT)) is a more efficient metric than RTT to identify wormhole attacks. Accordingly, we propose a new algorithm called Neighbor-Probe-Acknowledge (NPA) to detect wormhole attacks. Compared with existing works, NPA does not need time synchronization or extra hardware support. Moreover, it achieves higher detection rate and lower false alarm rate than the methods using RTT under different background traffic load conditions.
Large-scale wireless sensor networks (LSWSNs) consist of a large number of wireless sensors that have processing, wireless communication and information acquisition abilities. LSWSNs are promising techniques in many areas such as target detection and tracking, commercial management, intelligent family, military use, preventing forest fire loss, medical diagnostic, etc. In LSWSNs, maximizing the network lifetime of each and every sensor node separately will not lead to optimum power utilization for the complete networking. Maximizing the network lifetime under the energy ability and sensing radius constraints is an NP-hard problem. In this paper, a novel adaptive hybrid clone genetic algorithm (AHCGA) is given and utilized on optimal node scheduling problem. We first develop our objective function to maximize the network lifetime under many constraints. The resulting AHCGA approach integrates the merits of adaptive adjusting and niche selection while maintaining the contention for the genetic algorithm. To illustrate the effectiveness of AHCGA, simulations are executed for the optimal node scheduling problem, and performance comparisons are made with ant colony optimization (ACO) and cat swarm optimization (CSO). Simulation results demonstrate that the presented technique can achieve a longer network lifetime over ACO and CSO.
The burgeoning interest in developing Large Language Models (LLMs) with up to trillion parameters has been met with concerns regarding resource efficiency and practical expense, particularly given the immense cost of experimentation. This scenario underscores the importance of exploring the potential of Small Language Models (SLMs) as a resource-efficient alternative. In this context, we introduce MiniCPM, specifically the 1.2B and 2.4B non-embedding parameter variants, not only excel in their respective categories but also demonstrate capabilities on par with 7B-13B LLMs. While focusing on SLMs, our approach exhibits scalability in both model and data dimensions for future LLM research. Regarding model scaling, we employ extensive model wind tunnel experiments for stable and optimal scaling. For data scaling, we introduce a Warmup-Stable-Decay (WSD) learning rate scheduler (LRS), conducive to continuous training and domain adaptation. We present an in-depth analysis of the intriguing training dynamics that occurred in the WSD LRS. With WSD LRS, we are now able to efficiently study data-model scaling law without extensive retraining experiments on both axes of model and data, from which we derive the much higher compute optimal data-model ratio than Chinchilla Optimal. Additionally, we introduce MiniCPM family, including MiniCPM-DPO, MiniCPM-MoE and MiniCPM-128K, whose excellent performance further cementing MiniCPM's foundation in diverse SLM applications. MiniCPM models are available publicly at https://github.com/OpenBMB/MiniCPM .
Abstract Energy underground structures, such as energy shafts and tunnels, are a class of energy-saving and environmentally-friendly underground structures. They can be used as shallow geothermal energy harvesting systems for heat exchanges with the ground. Theoretical research and practical applications of energy shafts are still in the initial stages, and related tests are still relatively few. This study used the Shanghai Taihe energy shaft test section as the research background. The results of the thermal force response test of the energy shaft was numerically analyzed using the COMSOL software and compared with that of Beijing Qinghuayuan energy tunnel for applicability evaluation and analysis. The results show that the Shanghai strata energy shaft has a better heat transfer capacity than the Beijing strata energy tunnel owing to the different geological environments and is more economical. Simultaneously, the temperature stress generated in the heat transfer process of the Shanghai Strata energy shaft was smaller than that of the Beijing strata energy tunnel, which is safer. Therefore, underground energy structures can be widely applied to the Shanghai strata, which can significantly improve economic applicability while ensuring safety.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)