Router Technology Overview

In today's information society, people have increasing demands for data communication. As the core equipment of the IP network, the router has become a key technology in the current information industry.

What is a router

A router is a packet forwarding device that works at the third layer of the OSI reference model—the network layer. Routers forward data packets to achieve network interconnection. Although routers can support multiple protocols (such as TCP / IP, IPX / SPX, AppleTalk, etc.), most routers in China run TCP / IP protocol.

Routers usually connect two or more logical ports identified by IP subnets or point-to-point protocols, and have at least one physical port. The router determines the output port and the next hop address based on the network layer address in the received packet and the routing table maintained inside the router, and rewrites the link layer packet header to forward the packet.

Routers usually dynamically maintain routing tables to reflect the current network topology. Routers maintain routing tables by exchanging routing and link information with other routers on the network.

The router is the core device connected to the IP network.

Router classification

The current router classification methods are different. The various classification methods are related, but they are not completely consistent.

In terms of capabilities, routers can be divided into high-end routers and low-end routers. The division of manufacturers is not completely consistent. Generally, routers with a backplane switching capacity greater than 40G are called high-end routers, and routers with a backplane switching capacity of less than 40G are called low-end routers. Take Cisco, which has the largest market share, as an example. The 12000 series are high-end routers, and the series of routers below 7500 are low-end routers.

In terms of structure, routers can be divided into a modular structure and a non-modular structure. Generally, high-end routers have a modular structure, and low-end routers have a non-modular structure.

From the network location, routers can be divided into core routers and access routers. The core router is located in the center of the network, usually using high-end routers. Requires fast packet switching capability and high-speed network interface, usually a modular structure. The access router is located at the edge of the network and usually uses low-end routers. Requires relatively low-speed ports and strong access control capabilities.

In terms of function, routers can be divided into general routers and dedicated routers. Generally speaking, the router is a general-purpose router. Dedicated routers usually optimize router interfaces, hardware, etc. to achieve certain functions. For example, the access server is used to access dial-up users, enhancing PSTN interfaces and signaling capabilities; VPN routers enhance tunnel processing capabilities and hardware encryption; broadband access routers emphasize the number and types of broadband interfaces.

In terms of performance, routers can be divided into wire-speed routers and non-wire-speed routers. Generally, wire-speed routers are high-end routers that can forward packets at the media rate; low-end routers are non-wire-speed routers. But some new broadband access routers also have wire-speed forwarding capabilities.

There are many router classification methods, and with the development of router technology, more and more classification methods may appear.

Router function

Routers usually implement the following basic functions:

Implement Internet protocols such as IP, TCP, UDP, ICMP.

Connect to a network where two or more packets are exchanged. For each connected network, the functions required by the network are realized. These features include:

The IP data packet is encapsulated into or taken out of the link layer frame.

Send or receive IP datagrams according to the maximum packet size supported by the network. This size is the maximum transmission unit (MTU) of the network.

Convert the IP address to the link layer address of the corresponding network. For example, the IP address is converted into an Ethernet hardware address.

Implement flow control and error indication supported by the network.

Receive and forward data packets, and realize buffer management, congestion control and fairness processing in the process of sending and receiving.

When an error occurs, the error is recognized and an ICMP error and the necessary error message are generated.

Drop packets with a time-to-live (TTL) field of 0.

Fragment the data packet if necessary.

According to the routing table information, the next hop destination is selected for each IP packet.

Support at least one internal gateway protocol (IGP) to exchange routing information and reachability information with other routers in the same autonomous domain. Supports Exterior Gateway Protocol (EGP) to exchange topology information with other autonomous domains.

Provide network management and system support mechanisms, including storage / upload configuration, diagnosis, upgrade, status report, abnormal condition report and control, etc.

Router technology

Router software

The core technology in router technology is software technology. Routing software is one of the most complex software. Some routing software runs on UNIX operating systems, some routing software runs on embedded operating systems, and even some software is itself an operating system to improve efficiency. Cisco, the world's largest router manufacturer, once claimed to be a software company, which shows the importance of router software in router technology.

Router software generally implements other functions such as routing protocol functions, table lookup forwarding functions, management and maintenance, and so on. Due to the large scale of the Internet, the routing table running in the router on the Internet is very large, and may contain hundreds of thousands of routes. It can be imagined that the lookup table forwarding work is very heavy. In high-end routers, the above functions are usually implemented by ASIC chip hardware.

The high complexity of routing software is also reflected in high reliability, high availability and robustness. The function of implementing routing software is not complicated. In free sharing software, we can even get the source code of the routing protocol and data forwarding. But the difficulty is that the software needs to run efficiently and reliably 24 hours a day, 365 days a year.

In the development process of routers, routers can be quickly realized by purchasing commercial source code. But it is generally believed that the router software needs one or even two years to stabilize.

Programmable ASIC

ASIC chip is an application-specific integrated circuit, which is the core technology of the current router to achieve wire-speed forwarding of data. Programmable ASICs concentrate multiple functions on a single chip, which has the advantages of simple design, high reliability, and low power consumption. It can enable the device to obtain higher performance and lower cost.

Through the use of ASIC chips, device port density can also be increased. The port density using an ASIC chip is several times the port density when using a general-purpose chip.

The design of the programmable ASIC chip is the hardware guarantee of the current high-performance router implementation.

Router interface

The router interface is used to connect the router to the network, and can be divided into two types: a LAN interface and a WAN interface. The LAN interface mainly includes Ethernet (10M, 100M and 1000M Ethernet), token ring, token bus, FDDI and other network interfaces. WAN mainly includes E1 / T1, E3 / T3, DS3, universal serial port (can be converted to X.21 DTE / DCE, V.35 DTE / DCE, RS? 232 DTE / DCE, RS? 449 DTE / DCE, EIA530 DTE) ATM interface, POS interface and other network interfaces.

The current router interface technology is relatively mature. The difficulty lies in the design and manufacture of high-density interface boards and the realization of high-speed interfaces (greater than / equal to 2.5Gbps).

Routing Protocol

The implementation of router routing protocol is an important part of router software. Routing protocols are used to establish and maintain routing tables. The routing table is used to select the output port or next hop address for each IP packet. The open routing protocol mainly includes RIP / RIPv2, OSPF, IS-IS and BGP4.

RIP / RIPv2, OSPF, and IS-IS are intra-domain routing protocols and are generally used inside AS (autonomous system) to calculate and exchange network reachability messages within the AS. RIP / RIPv2 is a distance vector routing protocol, which is generally used in small-scale networks within enterprises. OSPF and IS-IS protocols are similar in principle and implementation, and are link-state protocols that are generally used in large-scale enterprise networks or operator networks.

The BGP4 protocol is based on the distance vector and is the only option for the current inter-AS routing protocol. Usually BGP exchanges a large number of network reachability messages, which is an important protocol on the IP network.

The implementation of routing protocols is similar to the requirements of router software, and it needs to achieve high reliability, high stability, robustness, and security. Router performance

Router performance usually includes the following:

Backplane capacity: Usually refers to the router backplane capacity or bus capacity.

Throughput: Refers to the router packet forwarding capability.

Packet loss rate: refers to the proportion of data packets that cannot be forwarded among the data packets that should be forwarded by the router due to lack of resources under a stable continuous load.

Forwarding delay: refers to the time interval between the last bit of the data packet to be forwarded entering the router port and the first bit of the data packet appearing on the port link.

Routing table capacity: refers to the number of routes that the router can accommodate during operation.

Reliability: Refers to indicators such as router availability, trouble-free working time, and failure recovery time.

Router queue management mechanism

Because the router is a packet-switching device and the bandwidth is statistically multiplexed on each port, the router must maintain one or more queues on the port, otherwise the router cannot process multiple packets and forward them to the same port at the same time and the QoS capability on the port And other issues. The quality of the queue management algorithm directly affects router performance, QoS capabilities, and congestion management capabilities. Usually queue management algorithms are divided into time-based algorithms, round-robin algorithms, and priority-based queues.

Time-based packet scheduling algorithms all have the same form. They maintain two time-scales for each group, one named start TIme-stamp and one named finish TIme-stamp. The router determines the next forwarded data packet based on the above time scale. The most common algorithms based on time scale are WFQ and WF2Q.

Another type of scheduling algorithm is based on the round-robin scheduling mechanism. Their working principle is similar to the multi-task round-robin scheduling in the operating system. Scheduling algorithms based on rotation usually have WRR, DRR and so on.

Priority-based queue management can schedule the forwarding of data packets in different queues according to the predetermined or user-specified priorities.

Routers usually use mechanisms such as RED (random early detection) and WRED (weighted random early detection) in the queue to avoid congestion.

MPLS technology

As an efficient IP backbone network technology platform, MPLS technology provides a flexible and extensible backbone network switching technology foundation for the next generation IP network. The use of MPLS technology can greatly improve the operating efficiency of the network, can realize the QoS division of IP online services, and can rationally allocate network resources through traffic engineering to achieve constrained routing. With these capabilities, the MPLS network will also be able to provide efficient VPN services and real-time services. It can be said that MPLS technology is likely to become a key technology in the evolution of IP networks to the next-generation carrier-grade IP networks. Therefore, MPLS technology may also be the key to whether routers become the core equipment of next-generation IP networks.

Although MPLS has various advantages, it has not been widely used on the Internet. The reason is that the protocol is not mature, there is a problem with interoperability among multiple vendors, and there is a problem with MPLS across AS or even across areas. VC Merge (VC merge) needs to be studied. However, in the present view, MPLS is the most ideal solution for implementing network-based VPN and can implement traffic engineering. Future IP network research must explore the possibility of using MPLS. Router equipment must consider implementing MPLS.

VPN technology

VPN refers to the establishment of a virtual private network on a public network. You can classify VPNs from different perspectives:

By access method

Dedicated line VPN: A VPN implementation solution for users who have access to ISP edge routers through dedicated lines.

Dial-up VPN (also known as VPDN): Refers to the VPN service provided to users who access ISP using dial-up PSTN or ISDN.

By agreement type

Layer 2 tunneling protocols: Point-to-point tunneling protocol (PPTP), layer 2 forwarding protocol (L2F), layer 2 tunneling protocol (L2TP).

Layer 3 tunneling protocols: Generic Routing Encapsulation Protocol (GRE), IP Security (IPSec).

The MPLS tunnel protocol can be seen as between layer 2 and layer 3.

Divided by VPN initiation method

Customer initiation (also called customer-based): The starting and ending points of VPN service provision are customer-oriented, and its internal technical composition, implementation, and management are visible to VPN customers.

Server-initiated (also known as customer-transparent or network-based): Install VPN software in the company's central department and ISP (called POP), customers do not need to install any special software.

According to the type of current carrier

Dial-up VPN service (VPDN): VPDN in the first division.

Virtual Leased Line (VLL): It is a simulation of the traditional leased line service. The IP line is used to simulate the leased line. Users at both ends of such a virtual leased line seem that the virtual leased line is equivalent to the previous leased line.

Virtual Private Routing Network (VPRN) services: There are two types. One is VPRN implemented using traditional VPN protocols such as IPSec and GRE. The other is MPLS VPN.

QoS on the router

The QoS on the router can be obtained by the following methods:

Obtained through large bandwidth. In addition to increasing the interface bandwidth, no additional work is done on the router to guarantee QoS.

Since data communication does not have a correspondingly recognized mathematical model as a guarantee, this method can only roughly use empirical values ​​for estimation. It is generally believed that when the bandwidth utilization reaches 50%, it should be expanded to ensure that the interface bandwidth utilization is less than 50%.

This is achieved through end-to-end bandwidth reservation. This method uses RSVP or similar protocols to reserve bandwidth end-to-end between nodes communicating within the entire network. This method can guarantee QoS, but the cost is too high, usually only runs on the corporate network or private network, and cannot be realized on the public network of the big network.

Obtained through access control, congestion control and DiffServ (Diff? Serv). This method cannot fully guarantee QoS. This can be used in conjunction with increasing the interface bandwidth to provide a relative CoS to a certain extent.

Obtained through MPLS traffic engineering.

Router security

The security of the router is divided into two aspects, one is the security of the router itself, and the other is the security of the data.

Because the router is the core of the Internet, it is a key device for network interconnection. Therefore, the security requirements of routers are higher than those of other devices. The security hole of the host computer makes the host inaccessible at most, and the security hole of the router may make the entire network inaccessible.

There may be management and technical reasons for the security vulnerabilities of routers. In terms of management, poor choices of router passwords, improper use of routing protocol authorization mechanisms, and incorrect routing configurations can all cause problems with router operation. Technically, router security vulnerabilities may have the following aspects:

vicious assault. Such as eavesdropping, traffic analysis, counterfeiting, retransmission, denial of service, unauthorized access to resources, interference, viruses and other attacks.

Software vulnerabilities. Backdoors, operating system vulnerabilities, database vulnerabilities, TCP / IP protocol vulnerabilities, network services, etc. may all have vulnerabilities.

The security of the data transmitted by the router can be provided by the network or provided by the user. If provided by the network, it is only relevant to the access router. The IPSec security channel can usually be provided by the access router to ensure security.

Specialed in developing and manufacturing in all kins of mobile phone chargers around 15 years  with our own brand MLF, also with much  OEM/ODM experience and including help our customer to do customized plastic housing mold . Our factory can output 300K pcs high quality USB 2.0 Mobile Phone Charger daily

Mobile Phone Charger

Mobile Phone Charger,Fast Phone Charger,Cell Phone Charger,Dual Usb Charger

Meile Group Limited ,