Advanced Methods For Complex Network Analysis

Protocol Analysis

Discuss about the Advanced Methods for Complex Network Analysis.

A communications protocol refers a system of rules that makes it possible for two or more communications systems entities to transmit information/ data through any kind of a physical quantity. They are the standards that define semantics, syntax, and synchronization of communication and ways of recovering errors. Software, hardware, or both of them can implement protocols. Protocols that are well defined are used by communication systems to exchange different messages with every single message having a precise meaning that is projected to occasion a response from a variety of probable pre-determined responses for the specific circumstances. The specific performance is usually not dependent upon how it is executed; protocols are developed into technical standards after all the concerned parties come to an agreement. Protocol analysis refers to the process is using the right software and/or hardware tools to obtan, decipher, then construe and respond to contents of data packet as they move through a network’s media. It entails the employment of esoteric software/ hardware tools to examine network traffic as it moves across the network.

To achieve protocol analysis, protocol analyzers are used; a protocol analyzer refers to tool that can either be a hardware or software utilized for capturing data traffic and signals over a communications channel. The channel can vary from being a local bus on a computer to a satellite link that offers a means for communication using standard networked or point to point communication protocols (Meghanathan, 2016). Every unique communication protocol has a specific different tool for collecting and analyzing data and signals. Analyzers types include telecoms network analyzers, network packet analyzers, IP load testers, and bus analyzers. Among the most popular and widely used analyzers include Wireshark which is the most popular and powerful network analyzer in the world. Another is the Network Analyzer Sniffer Tool (NAST) which is an n-curses based network analysis tool. Another popular analyzer is the Angry IP scanner which is a cross platform open source protocol analyzer (Wallen, 2013).

The analysis using the protocol analyzer should start at the client end; this is the easiest part to start from for slow connections. While it may be difficult to understand the problem until traffic is captured at the server end, this approach makes it easy to read the trace file if only a single client experience is captured; it is important that the performance problem is reproduced using the analyzer tool. The next step involves looking through the trace file to establish where DNS queries are initiated by the client for sluggish response to the server. This requires connecting the analyzer to the server side. It is possible that the server already in the DNS cache so a TCP SYN can be sent to the application server to ascertain the response of the application from the server. If there is a quick DNS response time, typically above 150 ms, a connection request should be sent to application server from the client. The connection should be filtered and isolated so as to compare the round trip time of the network with the server response time. The difference in time between the TCP SYN sent by the client and the TCP SYN-ACK the server sends back is analyzed to benchmark the connection setup time. Using the delta time column, the time taken for the server to respond to the request; with this information, it can be determined where next to troubleshoot; for instance, the problem will be on the server end if there are no TCP retransmissions and the server has a significantly higher response time (Greer, 2009).

Distance Vector and Link State Protocols

The distance vector link protocol and the link state protocols have some fundamental differences between them, and have mainly to do with the nature of the routing information routers send amongst themselves. For distance vector protocols, the neighbors receive a list of the whole known network well as its own distance to all the networks from the router. The distance vector protocol bases the choices they take on the best path that can be taken to reach a specific destination, centered on distance, which is measured in hops; every time a data packet transits through a router, it implies that a hop has traversed. The route having the smallest number of hops before reaching a given network is assumed to be the most suitable route for the network. An example of where the distance vector routing protocol is used in local area networks (LAN) that uses the RIP protocol (Antoniou, 2007). Given four routers A, B, C and D where data is tio be transmitted from A to B, with A to B having a slower speed of 128k ISDN and A-C, A-D, and D-B having higher 100b Tx speeds, the chosen route would still be A to B over the 128 k ISDN, which is ten times slower than the A C D B direct route as illustrated below;

A link state routing protocol creates a complete network topology picture and is also termed the shortest path first routing protocol. With the link state, three separate tables are created on each router that has the link state router enabled. One table holds details on neighbors that are directly connected, the other table holds information on the entire network topology, and the last table holds information on the actual routing table. Information is directly sent by the protocol to all routers that are connected within the network. The protocol operates by discovering its neighbors and builds its neighbors table and then measures the total delay to every neighbor. A routing adjustment is then constructed and sent communicating all that it has already learned all the network’s routers and then it applies a suitable algorithm (the Dijkstra algorithm) to develop the shortest possible path to all the potential destinations (Antoniou, 2007). An example of the link state routing is what happens in the Internet. Having the same network as above in link state routing, Given four routers A, B, C and D where data is to be transmitted from A to B, with A to B having a slower speed of 128k ISDN and A-C, A-D, and D-B having higher 100b Tx speeds, the chosen route will now be A C D B shortest path route as illustrated below;

VLAN Concept

The machines within the buildings are found within the LAN of that building and form an Ethernet network with each building having a switch and a router. With a router present, the addressing for the different computers within the LAN is done automatically, so no need for manual addressing. The addressing will work based on the IPV4 protocol. The computers in every building should be divided into a subnet to logically supervise addressing so that for each building, the computers are addressed using an identical, common, and most significant bit group for their addresses. This will allow for future expansion of addresses within each subnet (Sonderegger, 2009). The VLAN concept also ensures better security because basically, a computer on another subnet cannot access those in other VLANs, except when such access rights are granted. The Class C addressing will be suitable for computers in a VLAN where the computers are less that 254 ad ensures the efficient functioning of the VLAN while also allowing the preservation of IP address space; this is important for future expansion. The VLAN itself will have an IP address and eliminates the need for a physical layer for VLAN. The VLAN will enable for future expansion if more computers and addresses need t be added to the company’s network (‘Cisco,’ 2015).

QoS (quality of service) is an industry standard mechanism that is meant to ensure high class performance in serious applications. Using QoS mechanisms, current resources can be efficiently managed by the by network administrators in ensuring the required service levels with reactive over provisioning of networks or expanding the network. QoS is important in ensuring all traffic within a network is treated equally so that all traffic gets the best effort from the network (Froehlich, 2016). The QoS ensures that the requests of some users and applications are more serious than others so some data traffic are given preferential treatment. It therefore ensures balance and high performance of network; this is achieved by the QoS ensuring sufficient controlling latency and jitter as well as sufficient bandwidth; it also works to reduce data loss. QoS accords administrators greater control over the network resources to better manage the network from a business point of view rather than a technical one. It improves the user experience and reduces costs through efficient use of resources while ensuring time sensitive applications that are mission critical has the requisite resources (Park, 2011). IP precedence is a type of service in which a 3-bit field, which treats highly significance data packets as having greater than other packets; so if there is congestion in a router and is congested and some packets need to be discarded, the packets with the lowest priority are discarded first. Diffserve is concerned with the classification of packets while they enter a local network, which applies to flow of traffic (Carrel, Tittel, & Pyles, 2016). Five elements define the Flow; the destination IP, source IP address, source port, transfer protocol, and destination port. ECN is an IP extension) that allows notification of end –to-end network congestions; without dropping data packets being dropped. The Type of Service (ToS) is a six bit IPV4 header and DSCP (differentiated services code point); the ToS specifies the priorities of a datagram and requests a route for high throughput, low delay highly reliable services (Bruno & Kim, 2004).

References

Antoniou, S. (2007). Dynamic Routing Protocols: Distance Vector and Link State Protocols. Pluralsight.com. Retrieved 13 April 2017, from https://www.pluralsight.com/blog/it- ops/dynamic-routing-protocol

Bruno, A. A., Kim, J., & Bruno, A. A. (2004). CCDA self-study: CCDA exam certification guide. Indianapolis, IN: Cisco Press.

Carrell, J. L., In Tittel, E., & In Pyles, J. (2016). Guide to TCP/IP: IPv6 and IPv4.  'Cisco',. (2015). Configuring Isolated Private VLANs on Catalyst Switches. Cisco. Retrieved 13 April 2017, from https://www.cisco.com/c/en/us/support/docs/lan-switching/private-vlans- pvlans-promiscuous-isolated-community/40781-194.html

Froehlich, A. (2016). The Basics Of QoS. Network Computing. Retrieved 13 April 2017, from https://www.networkcomputing.com/networking/basics-qos/402199215

Greer, C. (2009). Identifying Slow Server Response at Packet Level (by Chris Greer). LoveMyTool - Building an Open Community for Network Management and Monitoring. Retrieved 13 April 2017, from https://www.lovemytool.com/blog/2009/01/chris_greer.html

Meghanathan, N. (2016). Advanced methods for complex network analysis (1st ed., p. 383).  Hershey: IGI.

Park, K. I. (2011). QOS in packet networks. New York: Springer.

Sonderegger, J. (2009). JUNOS high availability (1st ed.). Beijing [u.a.]: O'Reilly.

Wallen, J. (2013). Five free network analyzers worth any IT admin's time – TechRepublic.

TechRepublic. Retrieved 13 April 2017, from https://www.techrepublic.com/blog/five- apps/five-free-network-analyzers-worth-any-it-admins-time