基于贝叶斯网络的移动业务客户流失预测研究

准确的客户流失预测能够使移动业务针对特定用户需求和消费水平进行差异化服务.本文提出了一种基于贝叶斯网络的客户流失预测模型,根据案例分析的结果识别客广1离网倾向原因,寻找具有高离网倾向的客户群,通过分类的特征分析,得出客户流失规则,并验证模型的有效性.     

A Management Platform for Commercial Web Services

Web Services offer the promise of allowing BT, along with its customers, partners and suppliers, to be able to assemble distributed applications that span multiple systems and organisations. The commercial drivers for this are cost reduction, new revenue opportunities and a better service through the use of automated end-to-end business processes. To achieve these benefits, BT will require a fully managed internal Web Services infrastructure. At the same time, creating a managed platform for our customers' Web Services deployments provides us with a new business opportunity. Early experience suggests that 20% of the effort in creating a commercial Web Services application is expended in defining the commercial opportunity, the service interfaces and server implementations. The remaining 80% is spent on ensuring that the operational and deployment environment is adequate to support business-grade service levels. Managing that 80% of the problem for our customers is a business opportunity for BT. This paper describes the requirements for a management platform for Web Services and then goes on to describe some of the value added business services that this platform can support. This revised version was published online in July 2006 with corrections to the Cover Date.     

基于营配调贯通的用电需求侧分析

国网公司已全面实施营配调数据贯通,以此促进公司专业管理和服务水平的全面提升。本文主要叙述实现营销SG186系统、电网GIS平台之间数据贯通的需求侧用电需求研究和优质服务分析,并就重庆公司营配调贯通方式进行了分析和阐述,同时提出了用电侧需求分析的自动化算法,并结合当前信息化建设的实际提出了结合需求侧分析和优质服务的客户定制服务软件,从供电企业的需求侧管理分析和客户关注的用电优质服务两方面提出了未来供电企业的发展方向。     

LTE系统中基于负载的能量有效性调度策略

针对绿色通信中如何能在保证一定服务质量的前提下降低能量消耗这一关键问题,根据带宽交换能量的思想,综合考虑系统吞吐量和用户间公平性,提出了LTE 下行传输中基于负载的资源调度策略,通过降低频谱效率换得能量效率的提高,以达到节能的目的。仿真结果表明,所提的两个算法在满足用户速率要求的同时都能在很大程度上降低发射功率,且保证了用户间的公平性。     

电力企业客户细分模型研究

通过客户细分,可以帮助电网公司更为准确地掌握用电客户差异化服务的业务需求,协助提供符合实际需要的主动服务,最终取得提升客户用电体验满意度和降低电网公司服务成本的成效。从电网公司用电客户细分现状与应用需求出发,结合广东电网客户细分的各类应用场景,提出了基于客户价值区间的客户细分、K-Means聚类客户细分与基于决策树的客户细分三种细分模型,研究并创新性构建了细分模型从目标群体选择、细分模型层级管理到细分结果的深度分析总结三步客户细分机制,为电网公司个性化优质服务策略的制定奠定了基础。     

An optimal power-dispatching control system for the electrochemical process of zinc based on backpropagation and Hopfield neural networks

This paper describes an optimization problem to minimize the cost of power consumption for the electrochemical process of zinc (EPZ) depending on varying prices of electrical power. A series of conditional experiments was conducted to obtain enough data, which reflect the complex relationships among the factors influencing power consumption. Two backpropagation neural networks are used to build a process model that describes these relationships. An equivalent Hopfield neural network is constructed to solve this nonlinear optimization problem with technological constraints, a penalty function is introduced into the network energy function to meet the equality constraints, and inequality constraints are removed by altering the sigmoid function. An optimal power-dispatching control system (OPDCS) has been developed to provide an optimal power-dispatching scheme and keep the EPZ running economically. Since the OPDCS was put into service in a smeltery, the cost of power consumption has decreased significantly, and it also contributes to balancing the power grid load.     

Introducing Fairness-Efficiency Trade-off for Energy Savings in Wireless Cooperative Networks

The focus of this paper is on a wireless cooperative network architecture, where a group of users exploits short-range wireless links to share the costs of a cellular download. To maximize the efficiency of the communication system, an optimization of parameters such as download time, monetary cost, and energy consumption can be implemented. Following this approach different portions of data shall be assigned for download to the involved users, which will then cooperatively exchange the contents on the short-range link. However, the policy of task assignment to the user terminals has a direct influence on the payoff of the single users, raising fairness issues in real implementation scenarios. Focusing on the energy savings introduced by the wireless cooperative network, in this paper we address the fairness issue by relying on game theoretic bargaining solutions. These solutions, have intrinsic properties to nicely model the duality between fairness and efficiency in the performances. An optimal trade-off algorithm between efficiency and fairness is then introduced, allowing the service coordinator to select the most appropriate bargaining solution and energy savings allocation under different constraints on fairness.     

Cloud download system optimizing by job and notification scheduling

With expansion of cloud-based applications, it is necessarily for business service provider to deal with how to provision resources to server customers. In cloud download, the service provider faces great challenge on determining when to process job and when to notify the user to meet the conflicting objectives, i.e., the service profit maximization and user satisfaction maximization. An economic model was proposed to find job and notification scheduling strategies for making tradeoffs between user satisfaction and profit of the system, in this model, optimization problems are formulated as follows, profit maximization under a satisfaction target and then satisfaction optimization with a profit bound. By solving these optimization problems, the optimal scheduling strategies are obtained under various quality requirements of service. It is demonstrated that the proposed strategy can obtain better performance than the system without the intelligent scheduling strategy.     

Flower pollination algorithm‐based energy‐efficient stable clustering approach for WSNs

The advances in the size, cost of deployment, and user‐friendly interface of wireless sensor devices have given rise to many wireless sensor network (WSN) applications. WSNs need to use protocols for transmitting data samples from event regions to sink through minimum cost links. Clustering is a commonly used method of data aggregation in which nodes are organized into groups to reduce energy consumption. Nonetheless, cluster head (CH) has to bear an additional load in clustering protocols to organize different activities within the cluster. Proper CH selection and load balancing using efficient routing protocol is therefore a critical aspect for WSN's long‐term operation. In this paper, a threshold‐sensitive energy‐efficient cluster‐based routing protocol based on flower pollination algorithm (FPA) is proposed to extend the network's stability period. Using FPA, multihop communication between CHs and base station is used to achieve optimal link costs for load balancing distant CHs and energy minimization. Analysis and simulation results show that the proposed algorithm significantly outperforms competitive clustering algorithms in terms of energy consumption, stability period, and system lifetime.     

Resource allocation mechanism in the TWDM-PON and C-RAN joint architecture with hybrid energy supply

Aiming at the problems of low resource utilization rate,high energy consumption and poor user service quality in the existing virtualized Cloud Radio Access Network,an energy-aware virtualized resource allocation mechanism with hybrid energy supply was proposed.According to the energy sources and energy consumption of different network devices,energy arrival and energy consumption models were established.Furthermore,under the premise of guaranteeing the quality of user services,considering proportional fairness and energy consumption optimization,distributed algorithms based on asynchronous update were used to allocate resources and harvested energy for different types of virtual cloud radio access networks and user virtual base stations to effectively improve the energy efficiency of network.The simulation results show that the proposed resource allocation mechanism can reduce energy consumption while effectively reducing the latency and improving the throughput.     

数据挖掘技术及其在经营分析系统中的应用

为了在市场竞争中保持及发展客户、提升客户价值,运营商迫切需要了解客户的行为特征以及业务发展、收益、营销、服务等方面的运营情况,以便及时制定更加合理和高效的经营策略,更好地在竞争中掌握主动.于是,电信企业纷纷着手建立数据仓库和在线分析处理(OLAP)系统,通过对运营支撑系统多年积累的业务数据进行整合和分析,以获得有价值的分析结果.但是,OLAP侧重的是对历史数据的分析,回答的是"过去发生了什么和为什么会发生"的问题,而运营商在对过去和现在的经营状况进行分析掌握的同时,更加渴望对整个企业未来的发展状况做出比较完整、合理、准确的分析和预测,于是,数据挖掘技术由于能够发现潜在的知识以及能够回答"将要发生什么"的问题而被电信企业加以"重用".本文简要阐述了数据挖掘技术、主要模型及方法、数据挖掘的实施步骤,并结合具体的案例说明数据挖掘在电信行业经营分析系统中的应用.     

Self-service reservation in the fieldforce

On-line service reservation is becoming increasingly popular in the service industry. As services become more complex, requiring co-ordination among individual actions, the planning for making service appointments is normally conducted manually by the operators, often resulting in slow responses to customers. To tackle this problem, we have developed a real-time service reservation portal that takes an order from a user and works out appointment options, based on the availability of enterprise resources, for the user to book on-line. The portal is capable of dealing with complex workflow appointing, supporting interactive fine-tuning in individual appointments, respecting customer preferences and work policies, monitoring progress of workflow, informing the customer of status changes in job scheduling, dispatching and execution, as well as handling exceptions automatically. We have integrated the portal, code-named FieldReserve, with other components in the BT fieldforce optimisation suite (FOS) that covers functions of resource planning, scheduling and execution for the fieldforce to demonstrate the feasibility and potential benefits of this technology. Benefits demonstrated are to provide customers with instant response of service reservation, to provide volume control on reservation, to break away from one-size-fits-all lead times for service fulfilment, to provide customers with equal access to service provision, and to bring transparency of service progress to customers. Currently, we are actively engaging with businesses and have obtained strong support to downstream the key values of this technology in commercial contexts.     

Fiber-Based Broadband-Access Deployment in the United States

After more than 20 years of research and development, a combination of technological, regulatory, and competitive forces are finally bringing fiber-based broadband access to commercial fruition. The three main approaches-hybrid fiber coax, fiber to the cabinet, and fiber to the home-are each vying for dominance in the industry, and each has significant future potential to grow customers and increase bandwidth and associated-service offerings. Further technical advances and cost reductions will be adopted, eventually bringing performance levels and bandwidth to gigabits-per-second rates when user demand warrants while keeping service costs affordable     

Service Innovation Involved by Employees and Customers： From Conceptual Model to an IT Service Case

With the service industry and information technology developing, Service innovation occurs continually in business today. It is important for service firms to understand the service＇s business and process; customer knowledge including needs, expectations and experiences. However, employees are key factor to collect the customer＇s advice. At the same time, employees may organize customers for the new service. So, the integration of employees and customers will help service firm in the way of developing a new service and upgrading a service. The paper describes some models to explain the relation based on some literatures and give some advices for service firms. At last, we introduce an IT service case     

Power analysis and minimization techniques for embedded DSPsoftware

Power is becoming a critical constraint for designing embedded applications. Current power analysis techniques based on circuit-level or architectural-level simulation are either impractical or inaccurate to estimate the power cost for a given piece of application software. In this paper, an instruction-level power analysis model is developed for an embedded digital signal processor (DSP) based on physical current measurements. Significant points of difference have been observed between the software power model for this custom DSP processor and the power models that have been developed earlier for some general purpose commercial microprocessors. In particular, the effect of circuit state on the power cost of an instruction stream is more marked in the case of this DSP processor. In addition, the processor has special architectural features that allow dual memory accesses and packing of instructions into pairs. The energy reduction possible through the use of these features is studied. The on-chip Booth multiplier on the processor is a major source of energy consumption for DSP programs. A microarchitectural power model for the multiplier is developed and analyzed for further power minimization. In order to exploit all of the above effects, a scheduling technique based on the new instruction-level power model is proposed. Several example programs are provided to illustrate the effectiveness of this approach. Energy reductions varying from 26% to 73% have been observed. These energy savings are real and have been verified through physical measurement. It should be noted that the energy reduction essentially comes for free. It is obtained through software modification, and thus, entails no hardware overhead. In addition, there is no loss of performance since the running times of the modified programs either improve or remain unchanged     

Service Innovation Involved by Employees and Customers: From Conceptual Model to an IT Service Case

With the service industry and information technology developing, Service innovation occurs continually in business today. It is important for service firms to understand the service's business and process; customer knowledge including needs, expectations and experiences. However, employees are key factor to collect the customer's advice. At the same time, employees may organize customers for the new service. So, the integration of employees and customers will help service firm in the way of developing a new service and upgrading a service. The paper describes some models to explain the relation based on some literatures and give some advices for service firms. At last, we introduce an IT service case.     

Energy-efficient communication networks for improved global energy productivity

Recently, a lot of effort has been put into research and development of energy-efficient devices and systems and into optimization of the network infrastructure with regard to energy consumption. On the one hand, reducing total power consumption of global communication networks has become an imperative and an important step towards the future highly energy-efficient Internet. On the other hand, capacity and performance of the Internet must be continuously improved in order to meet the high requirements set by the continuously increasing of both the number of broadband subscribers and the amount of user traffic. Thus, only a highly energy-efficient and high-performance network infrastructure designed to efficiently support advanced applications and services for improving energy productivity in many other branches of business and society can significantly contribute to global energy savings. This paper briefly discusses potentials and options for achieving such an energy-efficient network infrastructure. Some examples of software applications and services for global energy savings are briefly reviewed. The aim of this paper is to draw attention to the need for a holistic approach to evaluate energy efficiency and to improve the global energy productivity through the use of high-performance and energy-efficient networks, services and applications.     

Impact of mobile instant messaging applications on signaling load and UE energy consumption

Mobile Instant Messaging (MIM) applications transmit not only user-triggered messages (UTMs), but also keep-alive messages (KAMs) via radio access network, which induces heavy burden in control plane channel and wastes user equipment (UE) energy consumption. In this paper, we deduce the joint distribution of KAM period and UTM mean interval from the MIM application traffic characteristics. Correlating the joint distribution with radio resource control (RRC) state machine in LTE networks, we derive two analytical expressions for the control plane signaling load and UE energy consumption respectively. Then, the variation of signaling load and energy usage is demonstrated with different settings of RRC release timer, KAM period and UTM mean interval. The analysis indicates that KAM period is the upper bound of RRC release timer when reducing the signaling load. Besides, five times of UTM mean interval is the upper bound of KAM period when reducing the UE energy consumption and signaling load. These results can guide both network operators and MIM application developers to properly set control parameters for balancing the signaling load and UE energy consumption.

     

Value optimization of survivable multilayer IP/MPLS-over-WSON networks

Network operators are willing to provide a range of services in the hope of maximizing their profits: from the highly available connectivity services for key business customers to the unprotected or even best effort services for residential customers. These services are being provided through IP/multiprotocol label switching (MPLS) over wavelength-switched optical networks (WSON) networks. Such multilayer network enables the application of optimal load balancing between the packet and the optical layer, optimizing both the cost of the packet layer and the utilization of the WSON. To provide highly available services, redundant network resources need to be added to the network providing survivability against failures; generally speaking, the higher the survivability degree, the higher both the capital and the operational expenditures (CAPEX and OPEX, respectively) of the network. In this work, we design networks to meet specific availability objectives considering single failures in optical links, IP/MPLS nodes, and optoelectronic ports. The benefits of the designed networks are evaluated from an economic perspective defining costs and revenues models and using Net Present Value as a metric to evaluate future cash flows after an investment. To this end, CAPEX and OPEX, including power consumption and maintenance, and penalties as a consequence of service level agreement breaches are considered. Exhaustive numerical results on several reference network scenarios demonstrate how the value of the network can be maximized by tuning availability objectives.     

基于异构数据的电力短期负荷大数据预测方案

随着多种可再生能源电力的接入,电力系统正在向更智能、更灵活、交互性更高的系统过渡。负荷预测,特别是针对单个电力客户的短期负荷预测在未来电网规划和运行中发挥着越来越重要的作用。提出了一个基于异构数据的电力短期负荷大数据预测方案,该方案收集来自智能电表和天气预报的数据,预处理后将其加载到非关系型数据库中进行存储并做进一步的异构数据处理;设计并实现了一个长短期记忆递归神经网络模型,用于确定负荷分布并预测未来24 h的住宅小区用电量;最后利用一个住宅小区的智能电表数据集对提出的短期负荷预测框架进行了测试,并使用均方根误差和平均绝对百分比误差两个指标,对比了预测模型与两种经典算法的性能,验证了所提模型的有效性。