Local Area Network Connections, Essay Example

Ethernets are considered to be one of the most common applications for Local Area Network connections or LAN’s.  The most widely used connections being 10baseT and 100baseT which is an indication of the data transfer speed measured in 10 mega bits per second. Average transmission speeds being up to 100 mega bits per second.

The diagram above  illustrates a typical Ethernet packing system. Different communication protocols use different conventions. With an Ethernet packet the start of the header and data elements are defined by their location applicable to the start of the package. It’s a bit similar to the concept of a letter. The header may be considered to be the envelope and the data area where all the information goes inside of the letter.

Frame Format of Ethernet Package

Ethernet breaks the data files into packages and these are different from other network configurations. Hence a frame may be regarded as a “package of information transmitted as a single unit ”  (HN Computing, 2007).  The frame can be from 64 to 1518 bytes long. Each frame will contain control information following the same logical configuration. The figure to the left provides an illustration of a sample ethernet frame configuration. ” the Ethernet II frame, used for Transmission Control Protocol/Internet Protocol (TCP/IP), which gets transmitted across the network”  (HN Computing, 2007), has become the main standard used for data transmission across networks including that of the internet.

Manchester 4B /5B

4B/5B encoding, within the context of Ethernets, is a block coding. Essentially a group of 4 bits is encoded allowing an extra 5th bit to be added. As the input data is absorbed 4 bits at a time, it allows for 16 different bit patterns. Manchester encoding came about as the need to address a bandwidth issue of the normal 20mhz signal. The more efficient coding optimized the bandwidth for improved transmission performance. This created the way for multi level transition signalling and as such 4B/5B encoding resulted in a 125mhz signal. Within Ethernet this facilitated further encoding using three signals instead of two ( -1, 0, +1) this reducing the bandwidth required to the physical cable to 31.5 MHz. ” Gigabit Ethernet utilises five levels and 8b/10b encoding, to provide even more efficient use of the limited cable bandwidth, sending 1 Gbps within 100 MHz of bandwidth.”  (Fairhurst, 2006).

This method of encoding has been criticised as not being very efficient in terms of optimizing transmission speeds. The method primarily used in Ethernet cabling is considered to be only 50% efficient i.e. ” to achieve a 100 Mbps rate with Manchester encoding, you need a 200 MHz clock speed.” (Network Computer & Program Resources, 2010).

8B /10B Encoding

In 8B/10B coding we are referring to the assignment of each byte of data being allocated to a 10 bit code. The byte is divided into 3 most significant and 5 least significant bits. The 10 bit code needs to contain a combination of : five ones and five zero’s, four ones and six zero’s, six ones and for zero’s. ” In Gigabit Ethernet this rate is then reduced using PAM-5 a 5-level code (achieving less bandwidth than possible with a 3 level code, such as MLT”.  (Fairhurst, 2006).    This method of encoding serves a number of distinct purposes. In the first instance it ensures that there are enough transmissions in the serial data stream in order that the clock may be future recovered from the embedded data. Secondly it maintains the d-c balance by transmitting the same number of zeros. An essential requirement because transporting data between several chassis may result in different ground or power supplies.  “For transmission, an 8B/10B encoding scheme is employed. When encoded for transmission, 8-bit characters are encoded into a 10-bit stream. On receipt, the arriving 10-bit coded word is decoded into an 8-bit byte.” (Grossman, 2001).

Cable grades for different speeds of Ethernet

Twisted pair Ethernet cabling is provided in different grades in order to permit faster speeds.  It comes in two options : (i) Shielded twisted pair and (ii) unshielded twisted pair.  The shielded is more expensive and most colleges and schools opt for the unshielded option.  each pair gets a different number of twists per inch of cable. The more twists the greater the transmission speed and the less interruption or interference from other nearby electrical devices. Another means of reducing interference is by the adoption of co-axial cabling. It contains a single copper conductor at the heart of the cable and this with a protective plastic shielding makes it highly resistant to signal interference or other external disruptions.

Another more popular cable is that of fibre optic cable that transmits light rather than radio wave signals. The centre glass core and protective shielding makes it work in a similar way to the co-axial cable. The advantage is that it has the ability to transmit data much faster and over greater distances.  There are two types of fibre optic cables available ” single mode and multimode. Multimode cable has a larger diameter; however, both cables provide high bandwidth at high speeds. Single mode can provide more distance, but it is more expensive.”  (University of South Florida, 2009).

Advantages and Disadvantages of Ethernet

Advantages

• The concept is relatively simple and easy to install.  Only T adaptors with no need for hubs, transceivers or other complex devices;
• Owing to the simplicity of design, Ethernet networks are often affordable and represent a good cost option;
• The co-axial cable ( the main choice) is well shielded and provides minimum service disruption

Disadvantages

• Difficult to change once installed.  Changing network configurations will result in some downtime whilst any new network configuration is installed
• Ethernet cables are very fault intolerant.  Hence if a device goes down on the network it will most likely impact the entire network
• Ethernet cables are very difficult to diagnose when problems occur.  It often has to be accomplished by a process of elimination and this can be extremely time consuming and expensive
• Co-axial cables are expensive and cannot be used in other network applications. This means they are redundant at time of network replacement

Ethernet Performance Issues

History has outlined that ethernet networks have a tendency to be susceptible to severe performance problems as the networks expand. Implementing layer 2 switching can help to improve performance.  Congestion often occurs for a number of different reasons but common faults are : too many hosts on a given segment, insufficient bandwidth, broadcast storms and too much congestion on the network.  The computer firm Cisco use the following types of metrics to assess an ethernet network performance.

• Network utilization. A network utilization of over 40% on shared Ethernet segments represents that the network is saturated.
• Broadcasts/multicasts. No Ethernet segment should have more than 20% combined broadcast and multicast traffic.
• CRC Errors. There should be less than 1 CRC error per MB of network traffic.
• Collisions. Less than 0.1% of network packets should be involved in collisions (Dinicolo, 2007)

Carrier Sense Multiple Access with Collision Detection

CSMA/CD defines the situation or event when two devices detect a clear channel and both try to transmit their signal simultaneously. This a collision occurs causing both devices to stop abruptly at the same time. Both wait for a random amount of time before retransmitting the signal. This technique is used in the 802.3 Ethernet network channel. ” It is estimated that network traffic must be less than 40 percent of the bus capacity for the network to operate efficiently. If distances are long, time lags occur that may result in inappropriate carrier sensing, and hence collisions.”  (Linktionary, 2009).

Works Cited

Dinicolo, B. (2007, 2 26). Ethernet Performance. Retrieved 8 9, 2010, from Ethernet: http://www.2000trainers.com/ethernet/ethernet-performance/

Fairhurst, G. (2006, 1 2). 4b/5b Encoding. Retrieved 8 9, 2010, from 4b/5b Encoding: http://www.erg.abdn.ac.uk/users/gorry/eg3567/phy-pages/4b5b.html

Grossman, S. (2001, 1 22). 8B/10B Coding. Retrieved 8 9, 2010, from Electronic design: http://electronicdesign.com/article/boards-modules-systems/8b-10b-encoding4276.aspx

HN Computing. (2007). Local Area Networks. Retrieved 8 9, 2010, from HN Computing: http://www.sqa.org.uk/e-learning/NetTechDC02CD/page_69.htm

Linktionary. (2009). CSMA/CD (Carrier Sense Multiple Access/Collision Detection). Retrieved 8 9, 2010, from Linktionary: http://www.linktionary.com/c/csma.html

Network Computer & Program Resources. (2010). Network – 4B/5B Encoding. Retrieved 8 9, 2010, from Network Computer & Program Resources: http://iptraf.org/network/page-4/4B/5B+Encoding.html

University of South Florida. (2009). Cabling. Retrieved 8 9, 2010, from University of South Florida: http://fcit.usf.edu/network/chap4/chap4.htm