Network analysis is a general term for graph theory analysis, optimization analysis and dynamic analysis of networks.Network analysis is to detect, analyze and diagnose all data transmitted in the network, help users eliminate network accidents, avoid security risks, improve network performance, and increase the value of network availability.Network analysis is the key part and the most important technology of network management.Network analysis generally includes the following analysis situations: quick search and exclusionNetwork failure； Find network bottlenecks to improve network performance;Find and solve various network abnormal crises to improve security;Manage resources, count and record the traffic and bandwidth of each node;Standardize the network, view various applications and services,hostTo monitor network activities;Analyze variousNetwork Protocol, manage the quality of network applications.
Network analysis is a method to calculate the network response when the excitation and network are known. It is also called circuit analysis.The most basic calculation rules are Kirchhoff's voltage law (KVL) and current law (KCL). In addition to the current voltage relationship (VCR) of each component in the network, enough network equations (usually calculus equations) can be obtained to obtain the required response.There are many different analysis methods according to the type of excitation source and network and the response to be solved.The method of directly solving the network calculus equation belongs to time domain analysis or time domain solution, where the excitation and response are functions of time t;The method of using Laplace transform or Fourier transform to solve the network equation belongs to frequency domain analysis or frequency domain solution, where the network equation is transformed into an algebraic equation, where the excitation and response are functions of complex frequency variables s or j ω.
Chinese name
network analysis
Foreign name
network analysis
Content
Graph Theory Analysis, Optimization and Dynamics Analysis
Applied discipline
network security、Administration、telecom
Role
Eliminate network accidents and avoid security risks
It is a common and important case that the excitation is a sinusoidal signal. The method to solve its steady-state response is the phasor method (also called symbol method).At this time, the excitation and response are sine functions of the same frequency, which can be expressed by phasors that only contain their amplitude and initial phase.For example:
Voltage phasor
express,
Current phasor
express.Using the phasor method, the differential equation can be transformed into an algebraic equation, and the current voltage relationship of network elements can be expressed by impedance or admittance.According to the different responses, there are many analysis methods, all of which derive their corresponding network equations on the basis of KVL, KCL and VCR.For a simple network, the network equation can be listed by observation method, and the solution process can be simplified by network theorem and equivalent transformation.For complex networks, it is often necessary to use network graph theory and matrix methods to systematically list the network equations and solve them with computers.The following six methods are commonly used.
2.1 Node voltage method
The voltage between each node in the network and a reference node is taken as the quantity to be calculated. This network equation is called the node equation, and its matrix formula is:
Where UnIs the vector of each node voltage to be solved;IgIs the current excitation source vector on each node is the node admittance matrix.
2.2 Loop current method
The imaginary current flowing in each independent circuit is taken as the quantity to be solved. The network equation at this time is called the circuit equation, and its matrix formula is:
Where ImIs the vector of the loop current to be solved;UgIs the voltage excitation source vector of each circuit;ZmIs the loop impedance matrix.
2.3 Port analysis method
Sometimes it is not required to calculate the current and voltage in the network, but only the current and voltage on the terminals connected to the network and the outside. At this time, the network can be treated as a multi terminal network. The most common network is a two port network. The process group connecting the current and voltage on these ports is generally small, which is easier to solve (two equations are only required for a two port network).
2.4 Network function method
When there is only one excitation source in the network (let its phasor be x) and only one response is required (let its phasor be x), the network equation connecting the two quantities can be derived as:
In the formula, H (j ω) is called network function, which is generally a function of frequency ω, and its dimension can be impedance, admittance, or dimensionless current ratio and voltage ratio, depending on the dimension of work and 7.Once H (j ω) is known, the response y can be calculated from the given x, and it is convenient to examine its frequency characteristics.
2.5 Indefinite admittance matrix method
Taking the voltage of the external terminal of the network to a reference point outside the network as the quantity to be calculated, the matrix formula of the network equation is:
Where, U is the vector of voltage between each external terminal and the reference point;I is the current vector of each terminal;YindIs the coefficient matrix of the equation, and is called the indefinite admittance matrix (singular matrix).Because it has a simple and systematic listing and solving method, and is suitable for computer processing, it is an important method for analyzing linear passive and active networks.
2.6 Topological analysis method
One kind of topological method is to express the relationship between the physical quantities such as current and voltage in the electrical network with a line graph, and then calculate the network function according to the simplified rules or formulas of the line graph. The typical method is the signal flow graph method.The other is to obtain the network function by calculating the tree branch admittance product of various trees according to the line diagram of the electrical network and the component parameters in the network.This method is called tree enumeration method or K-tree method.Topological analysis methods are suitable for computer processing and are easy to derive symbolic network functions, but they can deal with small scale electrical networks.
The DC excitation can be treated as a special case where the sinusoidal excitation ω is equal to zero. For periodic signals, it can be decomposed into many sinusoidal components with different frequencies by means of Fourier series. Since the linear network obeys the superposition theorem, the response of each sinusoidal component can be calculated by phasor method and then superimposed.
Linear network analysis under the excitation of aperiodic signals can be solved by means of Laplace transform, which converts the differential and integral equations of the network into algebraic equations, represents the current voltage relationship of network elements with operational impedance and operational admittance, and converts the sum in the network into complex transformation expressions V (s) and I (s).This method can be regarded as a generalization of the phasor method. It replaces j ω in the phasor method with the complex frequency s (here s=δ+j ω), so it is called the operation algorithm.It can be used in various solutions of phasor method.If the time domain function formula of the response is also required, the transformation formula of the response shall be obtained by Laplace inverse transformation.
3.1 State variable analysis
It can be used not only in linear time invariant networks, but also in time-varying and nonlinear networks.For linear time invariant networks, capacitance, voltage, inductance and current are usually taken as state variables, and a set of first-order differential equations state equations are derived with them as the variables to be solved.The state equation can be derived from the topological graph of the network, or from the higher-order differential equation or network function of the network.The advantage of this method is that there are abundant methods to solve the first order differential equations, and it is suitable for computer processing.In addition, it is also easy to apply time-varying networks and nonlinear networks.
In addition to the state variable method, the time-varying network function can also be used for linear time-varying network analysis.For nonlinear networks, the network equations derived from KVL, KCL and VCR are nonlinear equations, which generally have no closed solutions and are usually solved by numerical or graphical methods.
3.2 Computer aided analysis of network
With the development of computer technology, a general network analysis program appeared in the 1960s. It not only facilitates calculation, but also promotes the development of network theory.This kind of general network analysis program can be used for DC analysis, sinusoidal steady-state analysis, transient analysis, noise analysis, tolerance analysis and nonlinear network analysis.There are many methods used in the program, such as improved node method, state variable method and mixed analysis method, and sparse matrix technology is introduced to improve the efficiency of solving equations.
Network management
Announce
edit
Improve the overall level of domestic network management
Compared with overseas, domesticNetwork managementIt started late, and the management level of users is not as good as overseas.Kelai actively applies this technology to network fault resolutionnetwork performancePromotion andNetwork security protection, aiming to improve the domestic network management level, shorten the gap with foreign countries, help users realize full visualization of their networks, see the essence through network phenomena, and truly control their own networks.
Help network managers refine network management
Network Analysis TechnologyIt is the key to network management and the necessary technology for the network to enter the deep level management. The popularization of network analysis technology is also an inevitable trend. Kelai actively helps users build a deep understanding of this technology!The free version of the software is launched on the official website for the exchange and use of technology enthusiasts. At the same time, a large number of technical learning documents and video materials are targeted, and forces are organized to answer users' questions and share resources in relevant forums, aiming to promote the popularity of this technology in China, so that users can effectively access advanced network management technology.
Promote the exchange of network analysis technology
Kelai SoftwareActively hold various technical exchange activities. In addition to regular product and technical training, exchange activities with finance, telecommunications, energy, government and other industries promote the application of network analysis technology in the industry more widely and deeply. At the same time, the application of technology is more in line with the needs of practical work.
Management Key
Announce
edit
Network management faces great challenges
Today's network management needs network analysis technology more than ever!The network is constantly producing new key applications. The expansion of the scale and the complexity of the structure make users' management costs and maintenance costs for the network continuously increase. At the same time, viruses, attacks and network failures are also threatening the healthy development of the network all the time.Effectively ensure the continuous, efficient and secure development of the network, making network analysis technology the key to network management!
Breakthrough direction
Announce
edit
Expand network vision
Through the comprehensive monitoring and analysis of network data, we can obtain the network problems caused by various network application behaviors from the network bottom data, and quickly locate them, so as to better prevent the security strategy and manage the failure and performance more reasonably.
Fine network management
Network maintenance is becoming more and more complex. It is an inevitable trend for network management to be refined. Network analysis technology is a powerful guarantee for refined management by capturing data packets at the bottom of the network and decoding, analyzing and diagnosing them.
Perspective on network behavior
Only through the use of network analysis technology can its network behavior, information interaction, and data transmission process be fully visible, and reliable data support can be provided for troubleshooting, performance improvement, and network security problem solving.
Discover the communication records and monitoring of the whole network, and actively discover the network value space and potential network hazards.
Track and establish the association index of network communication data, high-precision and in-depth mining of massive data, and long-term trend statistics.
The reorganization, restoration and reproduction of forensics network events are used for security event forensics and network problem prevention.
network analysis
Announce
edit
The network analysis system is a network management solution that allows network managers to find the right solution to various network problems. It detects, analyzes, and diagnoses all data transmitted in the network, helps users eliminate network accidents, avoid security risks, improve network performance, and increase the value of network availability.
Managers no longer need to worry about the difficulty of solving network accidents. Kelai Network Analysis System can help enterprises minimize network failures and security risks, and gradually improve network performance.It brings network managers:
Quickly find and troubleshoot network faults;
Find network bottlenecks to improve network performance;
Find and solve various network abnormal crises to improve security;
Manage resources, count and record the traffic and bandwidth of each node;
Standardize the network, view various applications, services, host connections, and monitor network activities;
Manage network applications.
Network backtracking analysis
Announce
edit
The continuous, efficient and secure operation of the network is the basis for the normal operation of user business.This requires the network manager to be able to grasp the key indicators of business application operation at any time, find abnormalities in time and give early warning, so as to realize active operation and maintenance and active management;When a fault occurs, it can quickly and effectively locate the problem point, identify the responsibility and analyze the cause, thus reducing the fault time;Once the network receives an attack or a security event occurs, it needs to have a means and basis to effectively locate, analyze and obtain evidence.Kelai Network Backtracking Analysis System is just an intelligent and distributed network analysis platform to meet the user's constantly improving network fault, performance and security analysis needs.
The value of retrospective analysis:
Kelai Network Backtracking Analysis System is a high-performance hardware platform integrating network seven layer protocol analysis technology, high-performance data storage and intelligent data mining technology, and distributed data processing technology. It can provide users with irreplaceable value of other network management and security products.
Discovery - Real time network communication analysis and monitoring, active discovery of abnormal behavior and security failures in the network.
Tracking - intelligent data mining and communication analysis, quickly locate the root cause of the fault, and restore the network attack process.
Forensics - restoration and reproduction of original network communication data, used for packet level audit and forensics of security events.
