考虑负载均衡的系统保护通信专网路由规划方法

针对系统保护通信专网局部节点与链路业务负载过重的问题,提出一种考虑负载均衡的系统保护通信专网路由规划方法。首先,阐述了系统保护通信专网的概念,分析了局部节点和链路业务负载过重的问题;然后,构建了综合考虑业务特性和备份路径的负载均衡路由规划优化模型,实现了模型的负载均衡路由规划求解;最后,以某省系统保护通信专网光传送网(OTN)拓扑为仿真网络,采用PSCPlanner电力通信网规划平台与K条最短路径(KSP)算法进行了对比实验。仿真结果表明:该方法均衡了链路容量,降低了业务请求拒绝率,有效地分配了网络资源。     

电力光通信网中风险均衡路由算法研究

在大力发展智能电网的背景下,支撑电网运转的电力光通信网规模日趋庞大,其承载的业务更加多样化。然而电力光通信网的业务路由规划主要以最短路径算法为主,导致电力光通信网存在业务重要度分布不均衡,从而导致网络局部风险过高的问题。针对上述现状,文章采用深度强化学习技术,以网络业务风险均衡为目标,提出了基于强化学习的电力光通信网风险均衡路由算法。该算法考虑业务重要度分布情况、链路容量和链路光信噪比,实现了电力光通信网风险均衡化。文章选取某省电力通信子网验证方案的有效性,研究结果表明,该方法能够有效地降低电力光通信网风险均衡度,为电网的安全运行提供有力保障。     

光网络RWA中一种建立在Dijkstras和FJ算法的新模型

随着IP业务井喷式的爆发,现在的宽带容量越来越面临巨大的压力。网络容量必须跟上这一需求,而WDM技术恰好能利用光纤的巨大带宽,满足这一要求,并能极大降低传输数字信号的代价。WDM全光网络将得到极大的应用。RWA问题一直是WDM中的重要问题,因为选择是否恰当直接关系到网络的性能,介绍一种基于FJ思想的新思路,并利用他修改权重,他既能减少网络阻塞率,又能合理分配网络资源,起到优化网络的作用。     

基于复合量度的电力光纤多电路配置可靠性

能源互联网的一个本质特征是通过先进的通信网络实现能源供给与消费间的双向互动。以光纤通信为基础的高可靠通信网络优化配置以满足多业务的服务质量(QoS)需求成为亟需解决的关键问题。从电力光纤传输网络的可靠性入手,并考虑电力通信业务的多QoS传输需求,设计一种基于复合量度的电路配置优化方法。有别于传统最短路配置算法,新方法在一次配置过程中同时规划m条可行配置方案,并且通过分析路径集的带宽、时延、丢包率、衰耗和成本等不同量度和光路全程可靠性,最终获得满足可靠性和传输QoS需求的光路配置方案,并给出优先选择序列。多电路优化配置将有效提升电力通信的可靠性,也可为未来多路流量分摊提供电路配置保证。通过某城区35~500 kV电力通信站网实例仿真计算,新配置方法能够一次获取4条同时满足业务多QoS需求的传输电路,实现主备电路保护。对电力光纤传输网络的规划和运行优化,提高电力通信网的资源调配能力和可靠性提供参考。     

多网络共存下基于贝叶斯的垂直切换算法

针对目前的垂直切换技术普遍不能满足多网络下的切换问题,提出了一种改进的基于贝叶斯决策的垂直切换算法。首先根据接入终端网络的信号强度、网络负载、误码率和网络阻塞率以及用户业务对网络的实际需求偏好建立多条件相关的切换概率分布,计算出先验切换概率;然后通过贝叶斯决策算法计算出后验概率来进行判决比较,从而选取最优的接入网络。仿真结果表明,该算法不仅能有效地实现不同接入技术之间的垂直切换,从而减少不必要的切换,提高了切换成功的概率,而且还能在维持各网络平均负载达到30%~40%时减轻通信热点的阻塞情况,为用户提供更优的服务质量。     

移动边缘计算中服务功能链的自适应优化部署策略

为解决移动边缘计算中面向用户的服务功能链（Service Function Chain,SFC）部署成本开销过大、时延过长问题,提出了针对SFC的支出成本与时延联合自适应优化的部署策略。首先,在虚拟网络功能（Virtualized Network Function,VNF）节点选取阶段,考虑路径损耗这一无线信道衰落问题,根据有线用户与无线用户的位置情况,选择当前最佳节点以降低SFC的响应时延。其次,在服务节点配置阶段,根据用户请求处理的数据内容的新鲜度记录,自适应动态增加和删减相应的缓存,利用资源感知算法在保证数据传递可靠性的同时,减少服务节点的配置个数,降低配置开销。最后,在SFC部署阶段,利用基于KSP（K-shortest Paths）的功耗感知算法确定最佳节点映射排序与通信链路,在减少通信链路重映射的同时还能保证部署的SFC的低成本与低时延。实验仿真结果表明,相比于已有方案,该方法能够有效降低部署成本与时延,并能对不同用户的SFC部署做到自适应优化,提高了SFC的部署成功率。     

A novel grooming algorithm with the adaptive weight and load balancing for dynamic holding-time-aware traffic in optical networks

To maximize the resource utilization of optical networks, the dynamic traffic grooming, which could efficiently multiplex many low-speed services arriving dynamically onto high-capacity optical channels, has been studied extensively and used widely. However, the link weights in the existing research works can be improved since they do not adapt to the network status and load well. By exploiting the information on the holding times of the preexisting and new lightpaths, and the requested bandwidth of a user service, this paper proposes a grooming algorithm using Adaptively Weighted Links for Holding-Time-Aware (HTA) (abbreviated as AWL-HTA) traffic, especially in the setup process of new lightpath(s). Therefore, the proposed algorithm can not only establish a lightpath that uses network resource efficiently, but also achieve load balancing. In this paper, the key issues on the link weight assignment and procedure within the AWL-HTA are addressed in detail. Comprehensive simulation and experimental results show that the proposed algorithm has a much lower blocking ratio and latency than other existing algorithms.     

An efficient algorithm for the link allocation problem on ATM-basedpersonal communication networks

The personal communication network (PCN) is an emerging wireless network that promises many new services for the telecommunication industry. The proliferation of demands for extending wireless services to integrated services which supports the transmission of data and multimedia information has resulted in the need for broadband wireless systems that are able to provide service capabilities similar to those of wireline networks. The ATM cell-relay paradigm is one possible approach to provide broadband wireless transmission with PCNs using the ATM switching networks for interconnection of PCN cells. In an ATM-based PCN, the communication path between a pair of mobile terminals might be elongated due to the mobility of the terminals. The link allocation problem is that of allocating backbone links among ATM switches to reduce the effects of terminal mobility on the performance of ATM-based PCNs. Huang and Wang (1997) have shown that this problem is NP-complete. In this paper, we propose a new efficient heuristic algorithm for the link allocation problem. One novel feature of our algorithm is that we are able to derive sufficient conditions under which our algorithm is able to guarantee optimal solutions. Our empirical study shows that the average lengths of communication paths obtained by our algorithm are shorter than those obtained by Huang and Wang's algorithm. In addition, the number of successfully established paths obtained by our algorithm is significantly more than that obtained by the aforementioned     

弹性光网络中结合预测的多维感知RSA算法

弹性光网络(EON)中的传统路由频谱分配(RSA)算法多考虑路由跳数或频谱资源占用情况,缺乏时域与相邻链路的信息有效利用。提出一种结合预测的多维感知RSA算法,对持续时间已知业务的历史时间信息通过后向传播神经网络预测未来业务的时间信息,在路由时综合考虑时间、频谱和相邻链路资源占用程度。仿真结果表明:与传统RSA算法相比,多维感知RSA算法能有效降低带宽阻塞率。     

Research and Improvement of Wireless Sensor Network Secure Data Aggregation Protocol Based on SMART

The privacy-preserving of information is one of the most important problems to be solved in wireless sensor network (WSN). Privacy-preserving data aggregation is an effective way to protect security of data in WSNs. In order to deal with the problem of energy consumption of the SMART algorithm, we present a new dynamic slicing D-SMART algorithm which based on the importance degree of data. The proposed algorithm can decrease the communication overhead and energy consumption effectively while provide good performance in preserving privacy by the reasonable slicing based on the importance degree of the collected raw data. Simulation results show that the proposed D-SMART algorithm improve the aggregation accuracy, enhance the privacy-preserving, reduce the communication overhead to some extent, decrease the energy consumption of sensor node and prolong the network lifetime indirectly.     

A new ant colony-based routing algorithm with unidirectional link in UV mesh communication wireless network

Aiming at the unidirectional links coming from nodes with different transmitting power and the obstacle blocking in UV mesh wireless communication network and the traditional ant colony algorithm only supporting bidirectional links, a new ant colony based routing algorithm with unidirectional link in UV mesh communication wireless network is proposed. The simulation results show that the proposed algorithm can improve the overall network connectivity and the survivability by supporting the combination of unidirectional link and bidirectional link.     

基于强化学习的光传送网路由波长优化

针对光传送网中动态业务的路由和波长问题,提出一种基于强化学习的深度路由波长分配算法DeepRWA。算法基于软件定义网络架构,通过强化学习灵活地调整控制光传送网,实现光网络路由波长分配策略优化。针对路由选择问题,结合链路上的波长使用情况,使用A3C算法选择合适的路由,使得阻塞率最小;针对波长分配问题,使用首次命中算法选择波长。考虑阻塞率、资源利用率、策略熵、价值损失、运行时间及收敛速度等多个指标,利用14节点NSFNET网络拓扑仿真实验。结果表明:当信道中包含18个波长时,与传统KSP-FF算法相比,所提出的路由波长分配算法的阻塞率降低了0.06,资源利用率提高了0.02,但运行时间有增加;在波长数超过45以后,与传统KSP-FF算法相比,所提算法保持阻塞率和资源利用率的同时,运行时间开始降低;当信道中包含波长数为58时,与传统KSP-FF算法相比,所提算法运行时间减少了0.07 ms。由此可见,提出的算法使路由选择和波长分配得到了优化。     

弹性光网络双重故障下的自适应保护级别算法

为了降低弹性光网络中双链路故障保护下的业务及带宽阻塞率、均衡带宽资源分配,提出了一种双重故障下的自适应保护级别算法。该算法综合考虑传输距离、调制格式等因素,根据链路频谱资源使用状态动态地更新链路惩罚系数,选取最佳路径进行传输,同时根据请求类型及网络空闲资源状态自适应地选择保护方法,为业务提供最大限度的保护。仿真结果表明,在双链路故障下,算法在阻塞率方面取得了较好的性能,同时均衡了网络中各链路上的资源使用。     

Rate allocation strategies for energy-efficient multipath routing in Ad-hoc networks towards B3G

Wireless mobile Ad-hoc network is a special network that all nodes can self-organize and work together. It is flexible to form a network and extend the coverage area dynamically without infrastructure, so Ad-hoc network is envisioned as cornerstones of future generation networking technologies (B3G or 4G). However, the dynamic network topology makes the communication cost not only the energy of source/destination nodes, but also the relay nodes. Another problem of the Ad-hoc network is it is hard to provide a stable and persistent quality of service (QoS), which is strongly required by the beyond 3rd generation (B3G) system. In this article, the authors establish a scenario that contains B3G cellular base station and Ad-hoc mobile nodes, and propose two algorithms minimum incremental rate algorithm and power feed-back rate allocation algorithm in multipath routing. The algorithms can maintain a constant total transmission rate and bit error ratio (BER) to provide the QoS guarantee and reach the minimum power consumption of the relay nodes by adjusting the rate of each path in the multipath routing.     

A new solution to the K-shortest paths problem and its application in wavelength routed optical networks

In communication networks, traffic carried over long paths suffers from a higher call blocking probability (CBP) than those carried over short paths. This is a well-known fairness problem. Such a problem becomes more serious in wavelength-routed optical networks (WRONs) due to the wavelength continuity constraint. This paper aims to enhance the fairness characteristic in WRONs by the novel classified alternate routing (CAR) approach. As a foundation of this approach, the first K-shortest paths (KSP) between a designated source/destination node pair in a given network should be obtained simultaneously. This has been studied as the KSP problem in the literature. In this paper, a new loopless deviation (LD) algorithm for solving such a loopless KSP problem will be proposed. It outperforms existing solutions in terms of running time in real-life implementations. In order to measure the fairness characteristic, we first classify all connection requests into categories such as 1-hop traffic, 2-hop traffic, etc., according to the number of minimal hop count between the corresponding source/destination node pair. We then quantify the fairness characteristic by the ratio of the average CBP of each traffic category to that of 1-hop traffic such that CBP ratios with a value closer to 1 are preferred. We will show that such a measure criterion is more precise and robust than existing ones in the literature. Finally, numerical experiments will reveal that the CAR approach outperforms existing fairness enhancement methods when considering the comprehensive performance in terms of the balance between the fairness characteristic and the overall CBP feature.     

基于多智能体深度强化学习的智能电网光网络切片方案

为了提高光网络对大规模、差异化电力业务的资源分配能力,降低大规模业务的算法训练时间,提出了一种基于多智能体深度确定性策略梯度算法的智能电网光网络资源分配方案。该方案考虑大规模和差异化电力业务,将智能电网光网络建模成多智能体系统,以最大化电网公司收益为目标,建立了智能电网光核心网络切片模型,进行网络资源分配优化,并采用条件判断映射,简化了优化问题。同时,把不同业务部署到不同智能体中进行运算,以降低训练时间,满足网络实时性需求。仿真结果表明,该算法具有更大的奖励、更低的成本、时延和训练时间。     

An efficient dynamic distributed optical link establish method in intelligent all-optical networks

In this article, an efficient, fast, and dynamic distributed optical link setup method is proposed. Two kinds of novel optical fast link release mechanisms (Ahead and Timeout Release) are presented for an optical-link establishment. They can dramatically reduce the blocking probability. For dynamic routing detection, a new kind of dynamic weighted Dijkstra algorithm (DW-DA) along with dynamic optical link load balancing is described. A variable mutation and crossover rates of a genetic algorithm (VMCR-GA) is used for fast wavelength assignment with two-novel-wavelength assignment rules in a wavelength relationship graph, which can reduce the necessary wavelengths and link establishment time. Through simulation giving the blocking probability and the time for link establishment on several well-known networks, the effectiveness of this method has been verified. The blocking probability of the network can be reduced significantly below that of normal routing and wavelength assignment (RWA). Furthermore, the calculating time for reaching the minimum blocking probability can be reduced dramatically.

Dynamic multicast sharing protection algorithm based on fuzzy game in multi-domain optical network

Aiming at the problem of dynamic multicast service protection in multi-domain optical network, this paper proposes a dynamic multicast sharing protection algorithm based on fuzzy game in multi-domain optical network. The algorithm uses the minimum cost spanning tree strategy and fuzzy game theory. First, it virtualizes two planes to calculate the multicast tree and the multicast protection tree respectively. Then, it performs a fuzzy game to form a cooperative alliance to optimize the path composition of each multicast tree. Finally, it generates a pair of optimal multicast work tree and multicast protection tree for dynamic multicast services. The time complexity of the algorithm is O(k3m2n), where n represents the number of nodes in the networks, k represents the number of dynamic multicast requests, and m represents the number of destination nodes for each multicast request. The experimental results show that the proposed algorithm reduces significantly the blocking rate of dynamic multicast services, and improves the utilization of optical network resources within a certain number of dynamic multicast request ranges.     

一种快速动态分布式光链路建立方法

提出一种改进的波长分配算法,用于波长路由光网络中快速动态分布式光链路的建立。该算法采用提前释放和超时释放两种波长和链路释放机制,减少了网络拥塞,更适合波长路由网络的实际运行规则。通过在美国自然科学基金网(NSF)上的仿真计算,与以前的研究结果相比,该算法可以减少网络的阻塞率8％,同时缩短了光链路建立的时间。     

SMDP-based sleep policy for base stations in heterogeneous cellular networks

A dense heterogeneous cellular network can effectively increase the system capacity and enhance the network coverage. It is a key technology for the new generation of the mobile communication system. The dense deployment of small base stations not only improves the quality of network service, but also brings about a significant increase in network energy consumption. This paper mainly studies the energy efficiency optimization of the Macro-Femto heterogeneous cellular network. Considering the dynamic random changes of the access users in the network, the sleep process of the Femto Base Stations (FBSs) is modeled as a Semi-Markov Decision Process (SMDP) model in order to save the network energy consumption. And further, this paper gives the dynamic sleep algorithm of the FBS based on the value iteration. The simulation results show that the proposed SMDP-based adaptive sleep strategy of the FBS can effectively reduce the network energy consumption.