Category:
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z
0
1
2
3
4
5
6
7
8
9
Position:IcFull.com » IC Electronic information » Category: A

IC Electronic information

A wireless network for optimal allocation of TD Design

In Electronic Infomation Category: A | on April 18,2011

Abstract: The wireless network configuration parameters are managed wireless network optimization one of the most important. Large amount of data for network configuration parameters, error-prone, the characteristics of strong empirical, a case-based reasoning system configuration parameter management structure. This paper describes the function modules, the configuration parameters management system into the case-based reasoning. This line of reasoning to knowledge acquisition is simplified as the collection of empirical knowledge and LTC1522CMS8 datasheet and case base structure as a basis to build case for indexing and LTC1522CMS8 price and retrieval algorithms to improve the efficiency of the new problem solving for intelligent network optimization platform provides a new way.

0 Introduction

Network construction with the TD into the optimization stage, wireless network optimization is an important content of the work. Wireless network optimization has been run by a network of traffic data analysis, field test data collection, analysis of configuration parameters, hardware checks and LTC1522CMS8 suppliers and other means to identify the causes of network quality, and modify configuration parameters, network restructuring , device configuration adjustments, in support of a variety of business, and satisfy the QoS condition, get a good network capacity to meet the needs of certain wireless coverage.

Wireless network optimization is one of the most important tasks of the configuration parameters for wireless network management and analysis. Wireless network configuration parameters are divided into parametric programming data and network data, with the rapid development of the network, wireless network configuration parameters increasingly large amount of data, the network administrator simply individual or traditional network optimization software for simple management is clearly not enough, so the intelligent network optimization platform is particularly important.

This paper, the current development of TD network optimization, network optimization platform management module configuration parameters in the introduction of smart management module, and network optimization to work with other modules for network optimization data provide strong support and a reasonable optimization.

1 Network Configuration Parameter Management Overview

Configuration parameters are important data wireless network optimization, the parameters of the various equipment manufacturers to set very different, some of these differences reflected in the name of the range, but also to achieve some of the ideas embodied in the functional differences. At present, network equipment manufacturers not only one or two, through a unified platform focused on managing these complex parameters, is to improve the network optimization and efficiency of operation and maintenance of an important means.

Network optimization configuration parameter management is the most basic functions of the platform, the wireless access network OMC-R system sub-module, the network configuration parameters are divided into parametric programming data and network data, configuration parameters management should be able to fully presents the parameters of the wireless network configuration that can simultaneously manage and synchronize two sets of data to achieve support for the work of the network optimization and effective management of the system.

For TD such as large-scale wireless networks, even the most experienced experts in network optimization, management of configuration parameters is a very time-consuming and error-prone and difficult task. Network Optimization platform management subsystem configuration parameters in the design objectives:

1) configuration management, system design parameters should have the same time support multi-vendor, multi-format, multi-version of the data management capabilities, analytical individual manufacturers configuration parameters, configuration parameters in accordance with the BSC, cell, power control, switching, adjacent areas, TRX, the channel parameters such as classification preservation.

2) browsing through the configuration parameters, query the current configuration parameters, including parametric programming is network data query and data query.

3) configuration parameters to achieve compared to the whole network or local area network element parameters and the parameters of planning data for comparison. Display differences. According to user-defined parameters associated with the verification rules, verification of the consistency of parameters and found inconsistencies in the given tips. Threshold according to custom or default threshold, check the abnormal parameter values to give tips. Parameters of the verification results to the report presented in a way to help people network optimization to optimize the network data. Comparison process parameters shown in Figure 1.


Figure 1 Comparison

process parameters

2 network configuration parameters Intelligent Management System

Here are a network configuration parameters intelligent management of the system. The system configuration parameters can be compared to analyze changes in configuration parameters, save the parameter change to the fact that the library, the configuration parameters recommending treatment, provide reference for the network optimization. The system structure shown in Figure 2. Collected by the parameters of the system modules, intelligent calibration module, the parameters of comparison module, the intelligent management module (box section) and the interpreter 5 parts.


Figure 2 Network configuration parameters intelligent management system structure

2. 1 parameter collection module

Parameter collection module parameter collection functions, according to different network elements to be collected items of different parameters, timing, or manually extract the work of collecting, collecting objects should include all network elements, for the wired part. Such as office data, relay number, etc.; the wireless section, such as the number of base stations and carrier configuration, switching model.

Network element of the module in accordance with the acquisition parameters, the most fine-grained to a single network element, network element group can set the parameters collected to achieve the classification parameter extraction, and then to table documents submitted to the smart calibration module.

2.2 smart check module

Module receives from the smart calibration parameters table file collection module, its intelligent verification. Smart Check the main points Category 3, 1st class items can not be empty, some fields the primary key for the data can not be empty, when the entry is empty, given systematic verification; Category 2 as the standard format item , the field has a fixed format, when the input data format does not meet the field requirements prompting; No. 3, with some types of data items is optional, the data range of fixed, direct the class data is defined as a range, with an option for binding, while the data can be automatically matched. Intelligent calibration data appear to be effective in reducing the matching errors, while reducing the input workload. The following table header parsing for the MDF storage structure:

Typedef struct_SHEET_HEAD

{

Unsigned short usNetElement; / / data file belongs to the network element

Unsigned short usVersion; / / data version number

Char acTblName [RDBS_KNL_MAX_TBLNAME_LEN]; / / memory data file name

Unsigned short usFieldNum; / / memory data contained in a single record number of fields

Unsigned short usRecLen; / / memory, the effective length of a single record of data

Unsigned short usRecNum; / / data file contains the number of records

} SHEET_HEAD;

2.3 Parameter comparison module

Parameter comparison module receives information from the smart calibration module, according to predefined rules to file 2 copies of the data to compare all the tables, comparing items include: table, header information (the number of records, the word section number, record length, etc.), the overall comparison of the data, then the index key field is level for the corresponding parameters of comparison records, comparing the results obtained and stored. Comparative results for the smart management module into the rules of the form, the rules include:

1) rules of sign indicates the results of the rules of the serial number.

2) rule conditions that reference the value of the rule should have the condition. Parameter comparison results in a different field should correspond to a condition of entry, compared to the results should correspond to the number of conditions.

3) the fact that the rules of the library in the fact that the success of the experts handle the fact that in the past.

4) confidence factor that the credibility of the rules cited.

2. 4 Intelligent management module

Intelligent management module is the core of the system modules, in the absence of strong theoretical model, domain knowledge is not entirely intelligent and experienced management module into the case-based reasoning method can effectively solve the bottleneck of a series of complex knowledge problem. Knowledge bottleneck problem for the module, using case-based reasoning approach to simplify the process of knowledge acquisition, its solution with high efficiency, high quality, with a sustained ability to learn, can improve the intelligence of the system. Reasoning inference mechanism using case retrieval method can effectively improve the efficiency of reasoning.

Smart management module, including knowledge base, inference machine, explain the mechanism and case-based reasoning. Which constitute the knowledge base case (Case Library) includes facts and rules, you can gather from the network optimization experts directly into the case, you can successfully deal with the problem in the case of learning. The following mainly discusses case-based reasoning expert system module, which is mainly based on analogical reasoning, the outcome of the user easy to accept. Create a network configuration parameters based on intelligent management system of thought, part of the intelligence case-based reasoning module design shown in Figure 3.


part of Figure 3, case-based reasoning module of intelligent design By comparing the results

parameter analysis, the case represented as triples: parameter comparison results, problem analysis, problem solved.

Order to improve the search efficiency, the first comparative results based on parameter changes in the parameters of the field called search conditions were summarized searching * for the initial formation of a new case to retrieve the results, calculated on the basis of a preliminary search with solving this case and similarity new source case. Target case and the overall similarity of cases, said the source, such as type (1) below:


Where, T the target case, S represents the source case; n is the number of fields in the source case; Wi denotes the i-field similarity weight, value 1 / n or the experts weigh; F (Ti , Si), said the source case and target case of the first i-field similarity; Si as the source of the first i-field case, Ti represents the case of the first i-source field name field in case the same goal. Suppose Si,

range [a, b], Ti and Si, the similarity according to formula (2) calculation.


Used to measure the overall degree of acquaintance with the goal of the source case similar to the case level, the greater the overall similarity, indicating that the source case and target case more similar. In the initial search based on the calculation of the overall degree of acquaintance, can greatly reduce the source to compare the number of cases, thus greatly reducing the matching time.

2.4.1 Case study and preservation

When the system is used to solve new problems, if the source case base case and target case the similarity between the similarity to 1 or greater than the set threshold, can learn from previous experience, you can choose to maintain the original case base with new cases of the same or refresh. If the case library

cases and very different problems to be solved, then we need to solve the problem by experts. Then this case as a new case stored. In this way, the system gradually accumulated experience in solving related problems, so the reasoning ability of intelligent module continues to increase.

3 Conclusion

Wireless Network Optimization Based on the characteristics of configuration parameters, the case-based reasoning applied to configuration parameters management system, greatly simplifying the knowledge acquisition; of the past the results were re-used to improve the efficiency of solving new problems. Configuration parameters to manage the introduction of case-based reasoning system can greatly facilitate the application and network optimization platform development, network optimization provide a powerful tool, is conducive to the integration of network optimization and intelligent platform.

LTC1522CMS8 datasheetLTC1522CMS8 suppliersLTC1522CMS8 Price

All right reserved:icfull.com © 2010-2016 Certificate