Wireless Network Architecture Course Review

Categories: Network

Review, Pages 7 (1663 words)

Views

Introduction

As pursuing my Masters in Electrical and Electronics Engineering at The University of South Florida, with a concentration of Wireless Systems and RF. It was exciting and challenging at the same time because the course structure was different from that I pursued my bachelor's, from Mumbai University, India in Electronics Engineering. But my undergraduate has really created a perfect base for me, to study the subjects in-depth with practical exposure at the USF. I believe with my hard work and persistence efforts outcome of these learned courses will help me to gain the practical experience in potential wireless systems industry

Sequence 1: Wireless Systems

In this fast-paced world and emerging new technology, there is continuous development and deployment of new wireless systems due to increasing wireless traffic data.

Don't use plagiarized sources. Get your custom essay on

“ Wireless Network Architecture Course Review ”

Get custom paper

NEW! smart matching with writer

Even the current 5G technology faces many challenges like coverage and deployment, frequency bands, security, and privacy. This subject has provided me with extensive knowledge of multiplexing techniques like TDMA, CDMA, OFDMA which forms the building blocks of current advance techniques like NOMA, non-orthogonal multiple access used in 6G technology.

This course has really emphasized me with the knowledge of modeling, design, analysis, and optimization of these networks. I came to know about various wireless communication systems like WiMAX, LTE Network Architecture: SAE (System Architecture Evolution) and E-UTRAN (Evolved Universal Terrestrial Radio Access Network) -Evolution from 3G. I learned about the cellular telephony to the ubiquitous WIFI networks, to today's 4G LTE, advanced WIFI networks. IEEE 802.11 Access categories AC, IEEE 802.11 EDAC, HCCA TSPEC Procedure (Admission Control) and wireless application models, protocols, and performance issues in detail.

One of the questions from the assignment which shows actual how to calculate the carrier's channels that were a practical example of cellular capacity

Example: Consider a city with an area of 1,300 square miles that is covered by a cellular system using an N =7 reuse pattern. Suppose each cell has a radius of 4 miles and the city is allocated 40 MHz of spectrum with a channel bandwidth of 60 kHz. Assume that the system is designed for 2% blocking and the offered traffic per user is 0.03 Erlangs.

Solution: With the given data following questions can be solved:

a) The number of cells in the service area: Total area coverage is 1,300 square miles. Cell radius is 4 miles, and the area of the hexagon is known to be 2.5981 R^2, where R is considered as the radius of the cell hence each cell covers 2.5981·4^2= 41.57 square miles and 1300 square miles in the area of the city. The number of cells is Nc = 1300/41.57 = 31 cells.
b) The number of channels per cell C is given by formula 40 MHz is allocated spectrum and channel bandwidth as 60KHz the channel width reuse ratio=7 and , C = allocated spectrum / (channel width* frequency reuse factor) = [40*10^6 / (60*10^3 *7)] = 95 channels / cell.
c) The traffic intensity per cell is calculated from the Erlang charts with m =95 number of channels and Pb =0.02 (given 2%) to be 84 Erlangs/cell. By looking at the erlang table with 95 channels which are closer to 100 channels and 2% 0.02 approximately closer to Blocking Probability 0.01 hence it is 84 Erlangs/ cell
d) The maximum carried traffic = number of cells * traffic intensity per cell =31 * 84 = 2,604 Erlangs.
e) The total number of users: if the traffic per user is 0.03 Erlangs, then the total number of users = total
traffic/traffic per user = 2604/0.03 = 86,800 users
f) The maximum number of served mobiles: All channels occupied, is 95 channels * 31 cells = 2945 users, this is just 2945/86,800 = 0.034 or 3.4% of the total customer base.

In the above example, I learned about the traffic load, trunking, erlangs and how every term is interrelated to each other. If one trunk or channel in use 100% of all the time that would be equal to 1 Erlang = 3600 call-seconds/hour. But practically a trunk cannot be in use for the complete duration of time i.e. 100% hence it requires proper scheduling of calls. Reuse pattern allows us to use the allocated frequency spectrum efficiently as it is limited, it is limited by co-channel interference. The Offered traffic load ?, is the rate at which work arrives at the system is measured in Erlangs. Carried Traffic Load is the amount of traffic that is actually carried and used. Overflow traffic is the difference between the offered traffic and carried traffic. Blocking is defined as the ratio of the Overflow Traffic to the Offered Traffic Load. There are various traffic models erlang b, erlang c which describes how traffic overflow is described. If total traffic and traffic per user are given the total number of users can be easily determined, the number of cells can be determined the area of a hexagon and the region needs to be covered by cellular. In cellular concept specifically hexagon shape is used because it resembles more like a circle it gives the accurate partitioning of the area, it also has 6 equidistant co-channel neighbors and tight cellular packing of data.

Performance Management for Mobile Data Network, for a good QoS (quality of service) the wireless data should satisfy that traffic levels are network conditions are fairly "stable" over time. Static network management works well in the voice world. Some factors that affect the performance of mobile data networks are unpredictable traffic patterns, multiple bottlenecks, and severe congestion. All of this can be avoided by using Dynamic Network Management, mobile, and TCP/IP addressing.

I also learned the OSI model and functions of each and every layer and their potential importance of each layer in the network architecture. Connection- connectionless oriented services. In connectionless services, there is no specific path defined by a sender to send data hence, no guarantee of data is been received by the receiver also throughput rate is not assured, whereas in connection-oriented data it uses the defined path to send a data and thus it is guaranteed. In connection, the establishment of the connection is in static fashion example is Frame Relay, MPLS (multiprotocol label switching), ATM whereas in connectionless it is in a dynamic way and hence it uses the resources efficiently example is ethernet, IP. Emerging 5G, and Internet of Things (IoT), and advanced WIFI/802.11 networks I got an understanding of the foundation networking technologies that enable these emerging wireless systems. Studied about Wireless Application Models, Protocols, and Performance Issues. This course emphasizes conceptual knowledge on the modeling, analysis, design, and optimization of these networks within the unifying framework of resource allocation and mobility management.

Broadband Networks

This course made me familiar with the routing, switching, scheduling, QoS quality of services, theory terminologies. Term broadband defines high-speed internet access, mainly it provided me with an understanding of network technologies used in high-speed internet services along with the core infrastructure of a network system. Concepts of traffic engineering, statistical multiplexing reliability were studied. Demand for faster internet and mobile is been increased from the past couple of years improvements in current 5G technology are required to achieve the success of 6G technology in the near future.

By learning this subject, it made me in-depth details of various protocol architectures like ATM (Asynchronous Transfer Mode), MPLS (Multiprotocol Label Switching), DWDM (Dense Wavelength Division Multiplexing) this technique combines several wavelengths together into a common fiber. SONET (Synchronous Optical Network) avoids the failure of the link segment because it can switch the traffic coming from one end to another alternate path this is considered to be one of the best advantages. Circuit switching and packet switching.

Congestion is an evident problem, as it degrades the throughput of the system it can be avoided using 3 ways node (switch) level by Switch design, packet queuing, packet scheduling and dropping policies. The second method is network level by connection admission control (CAC) or Data path routing. The last method is application level by flow control or retransmission methods. Packet switch consists of a control card, line card and interconnection fabric. Line cards can be of two types ingress line cards its function is header processing, demultiplexing and egress line cards function involves multiplexing and scheduling. interconnection fabric (switching fabrics) is used to transfer packets between line cards.

Generic packet switch

The switching rate can be defined as the rate at which the packets are transferred from inputs to outputs. There are 3 ways of switching fabrics bus, crossbar and memory

Switching fabrics

I also learned about the Multi-Stage Switching design, some examples include the Banyan, Benes and Clos designs. Banyan and Benes are self-routing switches. These switches avoid ''All or nothing approach''. Banyan switch (8X8) shown in the below figure. This is the self-routing design and does not require controller, there is also other types of banyan design called batcher banyan it sorts the input based upon output ports to reduce collisions.

Banyan Design

Benes design is shown in the below diagram, it defines the specific route to reach one port to other port and can be designed in a various way by assembling several 2X2 crossbar switches this design can be implemented. Used widely in ATM switches and telephone networks

Benes Design

There are two basic types of switching circuit switching and packet switching. In circuit switching is connection-oriented implemented on the physical layer of the OSI model. It has a defined path between source and destination it is mainly used for voice communication by using space-division switching or time-division switching techniques. In packet switching, the path is dynamic not defined, packet switching is implemented in network level by using the virtual circuit approach or datagram approach. I learned about how queuing theory and how it is related to the Markov process. Broadband technologies should have optimum efficiency because the performance and success of emerging IP networks are dependent on it.

Project: Random Early Detection (RED) Algorithm

To avoid the severe problem of congestion in a network system this algorithm was used by and performed on the Linux operating system. In this algorithm prior to queue gets full the unwanted packets are dropped by calculating the statistical probability of each packet in a queue. This algorithm provides us with the benefits of global synchronization, reduced delay and fairness by reducing bias against the bursty traffic.