Internet Access Broadband

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

1.1 Internet Access

To say that Internet has become an important commodity with huge demand among the people would be an understatement. It has become an indispensable part of people's everyday lives. The Internet has simply shrunk the world into a global village, and brought everyone closer. We know there are few ways to access the Internet, namely:

Dial-up access - With this, Internet is accessed by making use of telephone lines via modem. This method is very slow and results in poor quality of the data. [7]

Broadband access - A broadband connection transmits data at a very high speed. With this video steaming and downloading can be done with great ease. It has now become a challenge for DSL, to provide the service as far as the remote and urban areas are concerned. [7]

WiFi access- WiFi is a well-known wireless technology which promises a great speed without any cabling requirement. A WiFi router can be set at home and internet surfing can be done in any corner of it. WiFi hot spots are available in various restaurant, libraries etc which makes it easy to do online stuff anywhere. [7]

All of the above said methods are struggling with one or more limitations. Broadband does not reach all the areas and also it is very expensive. WiFi, on the other hand, is mostly used indoors which results in a limited range of distance to access the Internet. So what can prove to be a long-term practical solution? A lot of researchers are seeing the answer to these questions in WiMAX.

Because of the emerging wireless technologies that enable users to move around or in other words, bring in the possibility of mobility, there is a need for a mobile IP.

1.2 Introduction to Mobile IP

With the Laptops, PDA's and other mobile devices becoming even more powerful and portable with the latest technologies , existing fixed networks need to meet up with this change and hence Mobile IP standard was proposed by Internet Engineering Task Force ( IETF ) in June 2006.

Mobile IP protocol allows user to retain same IP address without loss of connection while moving across multiple networks.

1.2.1 Need for Mobile IP Protocol [24]

Scenario: Imagine a traveler in a train downloading some files from the Internet on his PDA. Now when the connection is established with the service provider's network, his device is identified with unique IP address. Now as the train moves, the PDA due to loss in signal strength latches on to a stronger signal of other service provider who intern assigns him some other unique IP address on that network. Since the IP address has changed, all transport layer connections will break down and the downloading will be disrupted.

Mobile IP was primarily developed to cope up with this situation of dynamically varying IP address.

  • Mobile IP was developed with following objectives in mind
  • To enable host to stay connected to the Internet irrespective of their location
  • To enable host to be tracked without changing their IP address
  • Should require no changes to software of non-mobile hosts/routers
  • Free from geographical limitations
  • No modifications to IP addresses or IP address format
  • Should Support security

1.2.2 Mobile IPv4 Entities [25]

Mobile IPv4 consists of three components:

  • Mobile node
  • Home agent
  • Foreign agent.
  • Node C initiates communication with Mobile Node and sends packets to Mobile Node‘s home address
  • Home Agent intercepts packets and forwards them to the Mobile Node (proxy functionality)
  • Mobile Node replies directly to Node C

A node that moves from a sub-net to another sub-net is called a Mobile Node (MN) with its IP address as Home Address.

The sub-net, to which the Home Address belongs, is called the Home Network and the routing entity on this Home Network that does the job of forwarding packets to the Mobile Nodes is called the Home Agent.

When the mobile node moves to another sub-net, this new sub-net is called the Foreign Network. The routing entity receiving packets on behalf of the mobile node on the Foreign Network is called the Foreign Agent.

Because of the operation of these Mobile IP entities, no changes are needed in any other part of the network like routers or other systems such as DNS.

1.2.3 Mobile IPv4 Behavior [27][28]

The stages in the operation of the Mobile IP are:

  • Agent Discovery: On a Foreign Network process by which a Mobile Node discovers a Mobility Agent (Home Agent or Foreign Agent).
  • Registration: Process by which a Mobile Node registers itself on a Foreign Network with the Home Agent for Mobile IP Routing and Packet Delivery Services.
  • Routing and Packet Delivery: Packets routed from a Mobile Node to a Correspondent Host and back. Agent Discovery

When a Mobile Node enters into a network, it first determines the type of network by monitoring for a local broadcast message from a Home Agent or Foreign Agent i.e. whether it is on its home network or on a foreign network. This message is called an Agent Advertisement Registration

When a Mobile Node is detected by agent advertisement in a foreign network, it must register with the Foreign Agent.

Thus registration provides information to Foreign Agent of the existence of the Mobile node and also informs the Home Agent of the location and care of address of Mobile node.

The care-of address refers to an IP address local to the Foreign Network that the Mobile Node is currently visiting which is usually IP address of Foreign Agent.

For registration, the Mobile Node sends a Registration Request message (RRQ) to the Foreign Agent. The Foreign Agent interprets the message and forwards it to the Home Agent (as specified in the RRQ or dynamically assigned).

On receiving a valid RRQ, the Home Agent pairs the Mobile Node Home Address with the current Care-of Address. The Home Agent sends a Registration Reply (RRP) with a code indicating registration success to the Foreign Agent. The Foreign Agent relays the RRP to the Mobile Node.

Summary of registration process is as follows:

1. Mobile node requests forwarding service bysending registration request to the Foreign Agent (RRQ).

2. Foreign Agent relays this request to the home agent.

3. Home Agent accepts or denies the request and sends registration reply to the foreign agent (RRP).

4. Foreign agent relays this reply to Mobile node. Routing and Packet Delivery

The Home Agent upon successful pairing of the mobile node and the care of address sets up an IP tunnel between itself and the care of address mentioned in RRQ. The Foreign Agent then routes these packets to Mobile Node using a static route. De-Registration

In the event of Mobile node returning back to the home network, the Mobile Node informs the Home Agent of its presence on the home sub-net through de-registration process.

1.2.4 Introduction to Mobile IPv6 [29]

Before understanding the major differences between the two protocols we should know the procedure of Co-located Care-of Address in IPv4

The Mobile IP obtains care of address using two ways

  • Agent Discovery and Registration Process described above
  • As a co - located address obtained via DHCP

Mobile IP uses the later process when it is unable to find a Foreign Agent on the foreign network.

The major change proposed in the IPv6 protocol is to eliminate the foreign node for routing the packets between the Home Agent and the Mobile Node. This is implemented by Mobile node acquiring a co - located care of address of the visited sub -net all the time.

1.2.5 Advantages of using co - located care of address

  • Ipv4 reduced the demand for IP addresses by sharing the same care-of address amongst several mobile nodes with the help of foreign agent.
  • With IPv6, we have virtually unlimited addressing space and efficient auto-conuration mechanisms.
  • Movement detection like Agent Advertisements in MIPv4 is replaced by IPv6 mechanisms like neighbor discovery.

1.2.6 Comparison of IPv4 and IPv6 [30]

  • Mobile Ipv4 reduced the demand for IP addresses by sharing the same care-of address amongst several mobile nodes with the help of foreign agent for forwarding packets. It also supports the use of co-located care-of addresses (COA). In contrast, MIPv6 supports co-located COAs only.
  • Route optimization is enhancement feature available in MIPv4, whereas it is an integral part of the MIPv6 specification.
  • MIPv4 route optimization technique requires traffic to be tunneled between the Foreign Agent and the mobile node leading to routing delays. In MIPv6, as the source address is always the care-of address, packets are forwarded without tunneling.
  • Home Agent must be involved in the set-up of optimized routes in MIPv4. In the case of MIPv6, the mobile node can initiate an optimized route to a Foreign Agent quickly and efficiently without involvement of the Home Agent.
  • In MIPv4, a Care of Address is always obtained from a Foreign Agent or via DHCPv4. In contrast a COA can be obtained via IPv6 stateless or stateful address auto-conuration mechanisms in MIPv6.
  • MIPv4 provides its own security mechanisms, whereas MIPv6 employs the IPSEC protocol suite.

1.3 Introduction to Wireless Mesh Networks: [34]

Wireless mesh networks refers to wireless networks in which every node connects directly to other neighbor nodes. The topology in which they are connected is mesh in which every node is connected to every other node for redundancy path. Each of the nodes consists of mesh routers and mesh clients. Each of the nodes provides packet forwarding which may be for the next neighbor or to a remote wireless destination.

A mesh router consists of several interfaces with support for various wireless technologies. Wireless mesh routers with gateway/bridge allows interoperability with 802.11, Wimax, cellular and sensor networks.

The benefits of Wireless Mesh Networks: [28]

  • Network robustness: WMN topology provides excellent robustness by providing redundant paths for intermediate failed nodes.
  • Power: With the substrate nodes designed with very low power requirements, can work with solar, wind or hydro-electric power.
  • Integration: With mesh networks exposed on top of the buildings, they are built in with water resistant boxes and are generally noiseless and typically small boxes.
  • Scalability: In a wired technology, it is difficult to provide a scalable solution. With the implementation of WMN, it is highly scalable when the number of users increases.
  • Ad-hoc Network Support: WMN are self-cond devices which dynamically self-organize so that the nodes automatically establish and maintain mesh architecture among other nodes.

1.3.1 Wireless mesh networks are classified based on the following architectures.

  • Infra structure Meshing[35] :

Infrastructure Meshing forms a type of infrastructure for connecting wireless clients. A mesh router forms an intermediate router connecting various technologies. For a mesh router with a gateway support, helps in connecting to the internet. Conventional clients can communicate with the mesh router through an Ethernet interface. If the conventional clients use the wireless technology as the mesh routers, they can communicate directly with the mesh router.

If clients use different radio technologies then they have to communicate to the base stations that have common Ethernet connection to the mesh routers. In turn Mesh routers will communicate with other mesh routers connected with other radio technologies.

  • Client Meshing[37]

In client meshing clients communicate within the same type of technology. So a mesh router is not needed for communication. A packet destined to remote clients goes through several hops before reaching the final destination.

  • Hybrid Meshing [38]

Hybrid meshing combines both the Infrastructure meshing and client meshing. Mesh Clients have the flexibility to communicate either through the mesh clients or through mesh routers. Hybrid meshing is the most ideal implementation of Wireless Mesh networks. With the use of infrastructure meshing provides connectivity within different wireless technologies such as Wi-Fi, Sensor networks, Wimax etc and with the incorporation of Mesh clients provides improvement in routing and connectivity within single wireless technology or the domain.

1.3.2 Application Scenarios for Wireless Mesh networks: [39] [36]

  • Broadband Home Networking: [39][36]

In broadband home networking, we have various clients such as TV, Laptops, Desktops, Camcorders, camera, PDA, Refrigerator all using different wireless technologies. There are generally some dead spots and a wireless mesh network provides a solution to connect all the devices with different wireless technologies. below shows a detailed explanation of WMN implementation of the broadband home networking.

  • Community and neighborhood communication[39][36]

Imagine, you have a small community in your locality .So all members of the community want to communicate with other. Although you are neighbors, in order for you to communicate with your neighbor you have to access internet which significantly adds to paying to ISP's and significant delay. However, if you had to setup a wireless mesh network, you inter-communication is limited within your domain. You do not necessarily have to traverse the ISPs to communicate. All members would need to access Internet only when they need to communicate outside its WMN.

  • Enterprise Networking: [39][36]

With Enterprise building co-located, you may setup a WMN for inter-building communication. A Wi-Fi may be used as the backbone wireless technology; WMN provides a scalable solution to connect several clients. below shows 3 building networked through Wi-Fi with a mesh router.

  • Metropolitan Area networks: [39][36]

WMN provides a costless solution to remote and underdeveloped places to connect seamless network connectivity. A WMN provides a much higher data rate transmission than any cellular network which makes it fastest communication entity. WMN metropolitan networks spans at much larger radius than the home, community and enterprise networks. A WMN provides scalability compared to existing wired technologies as the deployment and scalable costs are much lower than the wired counterpart. The below shows a WMN implementation for a metropolitan area network.

1.3.3 Security of WMN: [40]

One of the challenges which WMN faces is its lack of secure communication due to its distributed architecture, nodes and intermediate routers susceptible to security attacks. Moreover wireless medium is till date provides a weak security owing to MAC spoofing etc. Attacker may modify the routing updates thus affecting the reliability of the routing protocols. The attacker may sneak into the wireless mesh network and impersonate another node. An attacker may spoof in the credentials of a particular node while authenticating or encrypting.

1.4 IEEE 802.11 Wireless LAN

Wireless LAN or WLAN allows two computers to connect to each other without any wires. It has attained tremendous importance due to its ease of installation, and with the gaining popularity of laptops. Public places such as Malls, cafeterias are attracting customers to use their premises for wireless access to lure the customers. For other‘s in educational and corporate world, it helps in the ease of communication and a way to miniaturize this world. Given the unprecedented usage, we are sure that wireless would eventually replace the wired world in the near future.

In a wireless network, devices communicate with each other via electromagnetic waves mainly with radio waves. Although radio waves travel at the speed of light, providing a great user experience of connecting to the Internet, they pose a serious problem of electromagnetic interference on account of its low frequency range. In order to reduce interference, we transform the frequency of the digital signals (usually in MHz range) into much higher frequency of the radio waves (usually several GHz) by a process known as encoding. Encoding greatly speeds up the data transfer and prevents “wasting” of the bandwidth. The two encoding techniques used in wireless networking are

Orthogonal Frequency Division Multiplexing (OFDM) [16]: OFDM divides the allocated frequency range into sub-ranges which simultaneously transmit the pieces of the data stream. The more channels we have, the more data can be transmitted in parallel, the greater bandwidth can be achieved.

Direct Sequence Spread Spectrum [17] (DSSS): DSSS increases the frequency of the digital signal by combining it with another signal of a higher frequency.

In order for the electromagnetic waves to carry digital data (0s and 1s), the waves need to be modulated. Among those used in wireless networking are: Binary Phase Shift Keying (BPSK), Quadrature Phase Shift Keying (QPSK), and Quadrature Amplitude Modulation (QAM).

IEEE 802.11[18] is a set of standards for wireless local area network (WLAN) computer communication, developed by the IEEE LAN/MAN Standards Committee (IEEE 802) in the 5 GHz and 2.4 GHz public spectrum bands. Below is the summary of the protocols used.





Frequency range (GHZ)



2.4- 2.4835


Encoding Method



OFDM ( DSSS for 802.11b compatibility)

Max Bandwidth

54 Mbps

11 Mbps

54 Mbps

Table 1 [19] [20][21][22]

Wireless technology has enabled internet to reach out to places that were inaccessible with a wired-infrastructure. Even in easily accessible regions, it provides a significant advantage of mobility, and reduced transmission cost. Understanding the wireless infrastructure is an integral part of the project. The wireless architecture consists of a system which is subdivided into several small cells (low power transceiver). These cells are grouped and each group is assigned a Base Station, also called Access Point (AP). There can be more than one access point in a domain, all connected to each other through a backbone network. A user joins the wireless network by providing a unique ID called the Service Set Identifier (SSID). The access point recognizes the SSID and authenticates the user. In order to provide mobility, a user is allowed to move from one cell to another and even from one domain to another.

Wireless networks still face many challenges compared to Wired Networks. Wireless networks still suffer from dead spots, i.e. regions without any reception at all. Even with very careful planning, some of the dead spots are hard to remove. Wireless environment raises additional challenges with its channel variability in time and space and user movement, limited frequency spectrum for transmission, interference among channels as frequencies have to be reused and, loss of privacy as air being the medium.

1.7 Introduction to Multimedia

Multimedia streaming has really matured as a technology over the last couple of years. It is giving an intense competition to the fixed internet. Few years ago when video/audio over internet had just started gaining popularity, users had to download media files to disks in order to listen-to/watch them. The streaming technology has relieved such pains. Other significant advantages of multimedia streaming include saving of terminal memory, easy distribution of large files, and no issues of illegal downloading and distributing copyright protected media content.

The basic idea of multimedia streaming is transporting the different types of desired media content to the user over a network, and displaying it to the user. Below, we give a brief overview of the components involved in streaming.

1.1 shows the three categories of multimedia streaming. ‘Stored audio/video' streaming involves a client sending a request to the server containing the desired content. The server responds to the request and transmits the content which is played black by the client's media player. Both, ‘Live and interactive audio/video' on the other hand, involve a client and two servers.

1.8 QOS

Earlier when IP was developed, the motivation was to have a communication between two entities. Applications during the earlier times were not susceptible to any delay and packet loss. The volume of traffic was also limited to a certain extent mainly due to the link capacity. But today's IP Networks are experiencing tremendous volumes of data. Moreover, the data is real time data. The high growth of data is influenced by increase in link capacity. The real time data is highly susceptible to packet loss and delay. So, the applications have a challenge in maintaining a good put.

QOS architectures were developed to reduce the delays for time sensitive data in the buffer and packet loss. QOS architectures are divided into two categories.

  • Integrated Services Approach (ISA)
  • Differentiated Services Approach( DSA )

1.8.1 Integrated Services Approach (ISA) [28] [41]

The Integrated Services Approach (ISA) is divided into two categories

  • Guaranteed Service Class: This class is pertaining to applications which are sensitive to delay and packet loss. It includes real time data such voice and video.
  • Controlled Load Service Class: This class is pertaining to applications which are less sensitive to delay and packet loss. It includes non-real time data.
  • Best effort service class: This class is pertaining to applications which do not need any guarantees of delivery.

Elements in ISA: [41]

  • Reservation Protocol: It helps to assign particular resources for each of the new flow.


  • Admission Control: It determines whether the new flow has sufficient resources before it can be serviced.
  • Management Agent: It helps to update the traffic control database and to assist admission control in order to provide admission control policies.
  • Routing Protocol: It consists of routing protocols such as EIGRP, RIP, and IGRP to provide the next hop address.
  • Classifier and Route Selection: A classifier sorts the incoming packets into appropriate buffers such as real time data. Based on the destination IP address and route selection is done.
  • Packet Scheduler: Packet scheduler schedules the appropriate packets depending upon the policy set for it.
  • Traffic shaping: Prevents bursts of traffic and helps in regulating bursty traffic.

1.8.2 Queuing Mechanisms: [41]

  • FIFO

In this type of queuing mechanism, the packets are scheduled in the first come first serve basis. In this queuing mechanism, there is no priority given to any specific packet. Another major concern is when the packets with smaller packet size are queued behind packets with larger packet size, the smaller packets experience more delay.

  • Fair Queuing

In this mechanism, the router puts all the incoming packets in pre-determined flows. Each of the flows functions as queue. Each of the packets from the queue is serviced in the round-robin fashion. It is named fair since every flow gets equal chance of sending the packets. But, a certain flow may experience its buffer getting filled faster than any other packet. So in this situation, packets may be dropped.

  • Weighted Fair Queuing

In WFQ, each flow is assigned a predefined weight. So, scheduler selects the certain flow for pre-defined period of time. WFQ is provides a certain uniformity of service to each flow.

  • Worst Case Weighted Fair Queuing (WF2Q+) :[42]

In WF2Q+, each flow is given a weight .The sum of all the weights is less than the predefined value W. Each weight associated with the flow tells how much it can share the link. In WF2Q+, the scheduler selects the next packet which is experiencing good flow whereas in the WFQ+ selects the next packet which had remained backlogged. So the packets destined to good receivers are unaffected.

1.8.3 Differentiated Services Approach (DSA) [28]

Diffserv approach provides a more scalable and simpler QoS architecture. It helps routers minimize their overhead by limiting the amount of storage.

The main block of Diffserv is its traffic conditioner. A traffic conditioner consists of the following components:

  • Traffic Meter: It measures the amount of traffic assigned to each traffic profile. It detects any violation of this rule.
  • Traffic Marker: It marks the packets in order to keep track of the packets.
  • Traffic shaper: It shapes the incoming packets to the outgoing port so that the receiver is in sync with the transmitter.
  • Traffic Dropper: It primarily drops packets which violates the traffic profiles.

Chapter 2


2.1 WiMAX Introduction

WiMAX (Worldwide Interoperability for Microwave Access) is an upcoming technology in the telecommunication industry, and it would be an understatement to say that it is growing at a fast pace. It is based on the framework of WMAN (Wireless Metropolitan Area Network), with an objective to offer a wireless expansion to cable and DSL for last-mile broadband access. The keyword ‘Last-Mile' basically refers to the concluding phase/stage of distributing the service of connection to a user from the service supplier. [8]

WiMAX standard is a combination of the two powerful Internet techniques namely broadband and wireless with no limitations for the Internet access. It is anticipated that it would provide services in those areas where wired connections is a big hassle.

WiMAX, also known as 802.16, can offer high-speed broadband connections over large distances. It can arguably be said that WiMAX is very similar to WiFi. However, unlike Wifi, it offers more speed, greater bandwidth and more secure encryption to multiple users. It operates at a bandwidth of 70Mbps covering a distance up to 10 miles. WiMAX uses the unregulated radio frequency spectrum, just like Wifi. But it does not need a line of sight and is open to multiple users per access point. Also, Wifi does not ensure any Quality of Service (QoS), but WiMAX promises to provide few stages of QoS. [5] It connects two points (stations) which are not in the line of sight and give an excellent performance over long distances. [1] More differences are covered in detail in later sections. WiMAX intends to makes use of a technology which consumes power at a low rate, and therefore, utilizes the Orthogonal Frequency Division Modulation (OFDM) technology. OFDM method is a transmission technique which is used to broadcast large amounts of digital data (audio and video) over high speed radio frequency and efficiently uses bandwidth. [1]

2.2 WiMAX Applications

WiMAX finds its usage in various applications like VoIP, streaming media, hotspots, mobile, also including last-mile broadband connections. [2] WiMAX would be available to users in two forms, mobile and fixed WiMAX. The fixed WiMAX form, known as 802.16d-2004, was considered to act as a replacement or appendage for DSL or broadband cable. A new version of WiMAX, known as 802.16e-2005, was recently approved to shore up the fixed WiMAX by allowing for roaming among the base stations. This version of WiMAX is called mobile WiMAX. Mobile WiMAX is based on OFDMA technology, which is another flavor of OFDM method, and assigns some sub-carriers to different users. It excels in performance by providing high throughput and efficiency. [1]

2.3 WiMAX Features:

WiMAX has the following features:

2.3.1 Flexible Structural Design: WiMAX is known to be very flexible as it provides a vast coverage and is also compatible with various conurations like Point-to-Point and Point-to-Multipoint. The MAC (Media Access Control) layer in WiMAX utilizes the Point-to-Multipoint protocol. [6]

2.3.2 Strong Security Mechanisms: WiMAX provides security by exercising various authentication and encryption methods. AES (Advanced Encryption Standard) and 3DES

(Triple DES) are the encryption methods which are being employed by WiMAX to protect the confidentiality of data transmitted between the base station and the subscriber station. WiMAX also guarantees the privacy of the users by not revealing any kind of information about the users in any circumstances. WiMAX also makes use of the EAP, which stands for Extensible Authentication Protocol. This method supports numerous authentication methods and assures advanced level of security. [6]

2.3.3 Speedy Layout: Deployment of WiMAX is not as complicated, lengthy and problematic as in case of the wired connections. It does not involve much of the outdoor set construction as digging the earth and putting cables. Also, the service providers who already attain the licenses to utilize the spectrum or who consider using the unlicensed bands, do not have to take more permission or do any extra paper work with Government.

Once the basic components like antenna and other equipment are set up, WiMAX is all set to rock. [6]In many situations, couple of hours is more than enough to set up and run the WiMAX services. [6]

2.3.4 Extensive Coverage: WiMAX incorporates a very good feature to provide extensive and widespread coverage of the service to the people. The reason is that WiMAX vigorously employs both levels of modulation:

  • High-level modulation: 16-QAM, 64-QAM
  • Low-level modulation: BPSK or QPSK

WiMAX system is proficient enough to cover a large part of an area if it works with a low level modulation and gets operational with a powerful amplifier

2.3.5 Tremendous Capability: WiMAX can provide a considerable amount of bandwidth to the customers if it exercises the higher modulation level and employs the channel bandwidth of 7 MHZ.

2.4 Components & Operational features

2.4.1 Base Station (BS) - A WiMAX Base station consists of the interior equipment and a WiMAX tower. The WiMAX tower is very similar to a cellular tower except that it can cover a radius of almost of 3,000 square miles. [5]

2.4.2 WiMAX receiver - The receiver could be a separate tower or a PCMCIA card inserted into your computer system. It is also known as customer premise equipment (CPE) or Subscriber Station (SS). With WiMAX, access to WiMAX base station is similar to WIFI, but here the coverage is supplementary. [5]

2.4.3 How WiMAX works: A high-level view - A base station can use the wired connection to gain the Internet access. It can also establish a connection to another base station using an LOS/microwave connection. A connection from an access point to the base station and further from the base station to the central (main) network is known as the backhaul link. As can be seen from the above, backhaul links can be utilized for connecting several base stations with each other.

This kind of mechanism enables WiMAX to provide the ‘roaming capability', just like roaming in cellular phones. [5] Users residing in buildings inside a range of 3-5 miles from a base station can establish a connection with the base station using the NLOS technology with excellent data rates. Users up to 30 miles away from the base station with an antenna mounted for line-of-sight (LOS) to the base station will be able to connect at data rates approaching 280Mbps.

WiMAX Physical and MAC layers are shown in the 1.

MAC Layer: WiMAX MAC layer is a point to multipoint protocol. WiMAX is a connection oriented Metropolitan area network. It supports difficult user environments which require high bandwidth and hundreds of users per channel. It utilizes spectrum efficiently by supporting both continuous and extremely bursting traffic. It grants high-quality service by setting up a specific time gap for each Subscriber Station (SS). In case if only one SS exists in the network, the communication between the Base Station (BS) and the SS will be done in a Point-to-Point form. In order to envelop a longer distances in its control the Base Station in a Point-to-Point pattern makes employs an antenna with a narrower beam. [6]

MAC Convergence Sub Layer: Offers support for ATM, Ethernet, 802.1Q, IPV4, IPV6 and there is possible future support for PPP, MPLS etc. The core MAC layer provides packet fragmentation, ARQ and QOS.

MAC Privacy Sub Layer: Integrates security features in WiMAX. Currently authentication, encryption and Key exchange functionality are provided in MAC sub layer.

WiMAX Physical Layer: This layer supports OFDM, DFS for transmission of raw bits. WiMAX physical layer employs the orthogonal frequency division multiplexing (OFDM) scheme. OFDM is a very a well-designed and competent format for transmitting data at a high rate in a non-line-of-sight (NLOS) situation.

2.6 WiMAX Service:

WiMAX offers two types of wireless services to the customers:

1) Line-of-Sight service (LOS): In this form, the antenna, on the roof or pole, is positioned in a way that it directly points at the WiMAX tower. This direct connection ensures stability and strength by transmitting huge amounts of data with less error.

LOS broadcasting technique makes use of high range frequencies as likely 66 GHz. The reason of exercising higher- wavelength transmissions that it guarantees much reduced amount of interference and provides extra bandwidth. [2]

2) Non-line-of-Sight (NLOS): This form of service is very similar to WiFi in the way where a tiny antenna on user's computer system connects to the WiMAX tower. NLOS guarantees improved coverage and uses a lower frequency range varying from 2 GHz to 11 GHz.

The reason for using the lower frequencies is that the physical obstructions like trees, walls, windows etc. do not interrupt them. The transmissions easily bend and diffract while they interact with obstacles on their way.

2.7 A Comparison between Wifi and WiMAX

WiFi and WiMAX technologies seem to go hand in hand. But they differ from one another in various aspects. Using an analogy, we can refer to WiFi as a cordless phone which can be used only within a restrict area like at home, rooftops etc. WiMAX, on the other hand can be thought over as a mobile phone, which is not restricted to be used in a bounded area. As cordless phones operate on frequencies, for which no license is needed to use, WiFi also uses the same frequencies commonly known as the unlicensed spectrum. For WiMAX, there are a few situations where unlicensed spectrum is used, but most of the usage would be in the licensed spectrum. Licensed spectrums are those radio frequencies whose use necessitates a license. [12]

WiFi is basically made for the LAN application. For each Consumer premises equipment (CPE) device, there is one subscriber. For each subscriber, there are around 1- 10 users. Also, here the channel size is fixed, normally of 20MHz. [5]

WiMAX, on the other hand, is being developed to manage hundreds of consumer premises equipments (CPE)'s, with no restriction to the number of subscribers associated with each CPE. It exercises the BWA (Bandwidth Wireless Access).The channel size is flexible varying from 1.5MHz to 20MHz.

If we keep the channel size as 20MHz, and look at the speed with which WiFi and WiMAX works, we are forced to take note of few interesting facts. With 20MHz channel size, WiFi works at 2.7 bps/Hz and can hit up to a maximum of 54 Mbps and WiMAX toils at 5 bps/Hz by reaching a maximum of up to 100 Mbps.

As there can be a huge amount of people accessing the WiMAX towers/base stations at one time, WiMAX should be able to provide excellent service to each and every user. Keeping that in mind, WiMAX is very much concerned about Quality of Service (QOS) and superbly addresses the issue. [14]

WiMAX forum is a non-profit organization which plays a very important role in sponsoring and attesting the IEEE 802.16 standard and equipment

2.8 WiMAX & QoS

Although WiMAX has the capability to deliver broadband speeds over 70Mbps but in order to provide good user experience, we need to be more concerned about the QOS design. [13] As mentioned above, WiMAX has applications in technologies involving streaming data like VoIP and IPTV. QoS is a very critical parameter to these technologies, and thus automatically becomes a very important feature in WiMAX communication. Providing efficient and guaranteed QoS is a challenging task especially when the services have to be provided to large masses. An ongoing practice for QoS provision in the multimedia industry is ‘SLA agreement'. An SLA agreement is a contract between the user and the service provider, which mentions some major QoS metrics. [4]

QoS in WiMAX is exercised by the lower two layers, i.e. Physical layer and the MAC layer. WiMAX has an embedded QoS mechanism that, by itself, makes it quite different from other wireless technologies. The mechanism requires establishing connections between the Base Station and the Subscriber Station (CPE device). Each packet type is serviced based on specific scheduling algorithms, which means that QoS parameters can be guaranteed for each flow. The idea is to allocate bandwidth to a specific ‘High-Priority' application without affecting other applications in the network. Thus, it promises low latency, data prioritization and guarantees bandwidth for ‘mission-critical' packets like voice, video etc. [3]

Chapter 3


3.1 Types of Buffer

  • Real Time Packets (RTP): This buffer incorporates real time packets of variable size .this type of packets needs a guaranteed and timely service. If the packets in this buffer for a particular CID are not served within the given time frame will result in appreciable loss of data. Our intelligent classifier decides on a particular packet if it can be serviced in a particular time. So, if a particular CID has bad channel conditions it would only admit real time packets with those CIDS that can be serviced in the given time frame.
  • Best Effort Packets (BEP): Best Effort Packets are those packets which do need any guaranteed timely delivery. They are less sensitive to any delay. Our Intelligent classifier admits packets in the Best Effort Buffer based on first come fist serve basis and services in that order.
  • Non-real time Data: These types of packets are those which are delay tolerant and generally have variable size packets.

3.2 Connection Identifier (CID): CID (Connection Identifier) is an integer that is unique for every Base Station connection with a Sub-Station (SS). The compensation block (covered in the next section), identifies each SS with this unique CID. Depending upon the CINR reports that the compensation block receives; the CID is classified as “good CID” or “bad CID”. This is also important for the classifier as our classifier, in case of real time packets, admits only a certain amount of bad CIDS. Classifier rejects packets for those CIDS which are experiencing bad channel conditions.

The above mentioned elements are the basic building blocks of the WiMAX QoS Architecture. In [44], authors have developed a ‘Channel-Aware' QoS architecture, which promises excellent QoS even when channel conditions deteriorate. The paper adds few more elements to the architecture (described in sections 3.3, 3.4 and 3.5) in order to make it ‘Channel Aware':

3.3 Compensation Block: A Compensation block's task essentially is to monitor the channel conditions for each Substation (SS). It is responsible for receiving CINR reports from Sub-stations that is utilized by the scheduler and classifier. In wireless conditions each substation periodically experiences varying channel conditions. Compensation Block is a key feature in the MAC layer architecture in order to provide a guaranteed channel aware QoS. A compensation block has to continuously monitor channel characteristics of each Substation. Depending on the channel characteristics, a CID is marked as ‘bad' or ‘good'. The principal of the model is that a good CID is to be given priority over a bad CID. When a Substation starts receiving good channel conditions, the service rate adjusts and it gets better service rate. So a compensation block is highly complex in architecture for a wireless system.

3.4 CINR Reports: The Base Station (BS) QoS architecture proposed in this paper can be improved to provide better QoS for delay -sensitive multimedia packets. The existing design makes use of a compensation block that receives Carrier to Interference Noise reports (CINR), also called bandwidth requests, from the Substations (SSs). For simplicity, in our project simulations we have sent only two kinds of CINR reports from the Substation: Good - signifying good channel conditions, and Bad - signifying bad channel conditions.

3.5 Scheduler: Compensation block connects to a packet scheduler that is supposed to select the appropriate flow for transmission. The scheduler works on the principal of Worst-Case Weighted-Fair queuing technique explained in section 1.8.2 of Chapter 1.

Before transmitting a packet, the scheduler checks whether the packet has a CID that has been marked as ‘banned'. If so, it looks to transmit a different packet with an unmarked CID, and the packet with the banned CID stays in the buffer. The scheme aims at conserving bandwidth by infusing intelligence at the scheduler level, thus, making the system channel aware.

3.7 Proposed Architecture

A major shortcoming, which also becomes the biggest flaw in the existing Channel-Aware architecture, is that the compensation block communicates only with the scheduler. The scheduler checks a CID's CINR report from the compensation block, before servicing a packet belonging to that CID. However, the classifier is completely unaware of these CINR reports, and continues to accept all packets as it is doing already. This process leads to accumulation of Banned CID packets in the buffer in case of bad-channel condition experienced by a Substation. Ultimately, the whole process enters a situation when a bad Substation's clogged traffic starts affecting other Substation's traffic, and increases queuing delay for packets that belong to Substations that were always experiencing good Channel conditions.

We propose an architecture that aims to avoid such situations, and stands strong in the worst channel condition situations. This architecture offers a better and improved overall QoS experience. We provide a feedback mechanism from the compensation block to the classifier, thus creating communication between the Classifier, Compensation block and the Scheduler. By providing this feedback, the classifier is given intelligence and is made aware of the channel conditions for each of the CIDs. An intelligent classifier is described in detail in the following section. The proposed architecture is shown in the below:

3.6 Intelligent Classifier: A classifier receives all types of packets randomly. It classifies the incoming packets based on the type i.e. real time, non-real time and Best effort packets.

In our proposed architecture, since the classifier is fully aware of the channel conditions, it admits in its buffers only those CIDs which are experiencing good channel conditions. It accepts packets with Bad CID up to certain threshold value, after which it stops accepting them, thus avoiding a situation when good packets start experiencing a significant amount of delay. Therefore, in our model, we implement ‘Admission-Control' at the classifier level itself, and accept only a pre-defined number of bad channel CIDs so that the buffer is not completely filled with bad packets. However, buffer for Best Effort packets is filled as normal as these packets are not really delay sensitive.

Another issue we are trying to address is the integration of an efficient buffer management scheme with a classifier itself. So classifier with a good buffer management scheme only admits real time packets which it knows it can service in the given amount of time, and also can eventually throw out the real-time packets that lose significance after some time.

The flow:

Let us assume that there are 5 Substations. Each substation sends either of the two types of reports i.e. Good OR Bad to the compensation block, based on the channel conditions. A feedback is provided to the scheduler and the classifier from the compensation block.

Each of the SS's periodically sends these reports. Based on this report and packet type, the classifier makes a decision as to whether to admit the packet in the buffer or not. A classifier makes decision on the following basis:

  • It admits all packets with ‘unmarked' CIDS.
  • If the report is bad, i.e. if the CID is marked as banned, it shall only accept the real-time packets with that particular CID up to a certain threshold level in the buffer, after which it would start rejecting the real time packets. Moreover, it shall also limit the amount of other types of packets depending on the number of packets in the buffer.

Next, the scheduler uses WF2Q method to schedule the packets. So when a particular flow is experiencing deteriorating channel conditions, it checks for other packets in the same buffer, as it improves service to packets with other CIDS, and conserves bandwidth at the same time.

However, the current implementation has a dumb classifier that is unaware of a particular CID marked as banned, and continues to accept packets based on the status of the buffers. Thus the buffers continue to get filled-in even though some of the CIDs that they contain are banned. Consider a scenario in which we say any one buffer is completely filled. This particular buffer consists of packets which belong to banned CIDS and unmarked CIDS. Although a fresh packet with a good CID comes in, the classifier rejects it since the buffer is completely filled, thus hampering the overall performance of the system.

In order to prevent such situations, we propose a design with intelligence at the classifier. So the classifier is informed by the compensation block of the channel conditions for each CID. With this help, Classifier can efficiently manage admission control. In the next chapter we discuss various scenarios that we have simulated to prove the effectiveness of the proposed architecture.

Chapter 4

Simulation Results and Analysis

In this chapter, we cover the scenarios that we have simulated and compare the results that we got with both the architectures.

In the following sections, we first describe the scenarios that we simulated and display the respective results:

4.1 Scenario 1 - Buffer Overflow:

This is the simplest scenario in which the base station receives a random order of packets until a point when it is attacked with a set of 200 Banned CID Real-Time packets and does not have a large enough Real-time buffer to hold them. After this set of 200 bad packets, we send a set of 50 good packets. The Objective was to compare Packet Loss. Plots, displaying the state of the Real-Time buffer, for both the architectures are shown below:

As can be seen from 4-a, after the Existing BS architecture has received about 200 packets, it receives a burst of 200 banned CID packets that the classifier admits, but the scheduler refuses to service. As a result, the buffer gets completely filled after sometime, and is unable to accommodate anymore packets. All the good 50 packets that were sent later get rejected by the classifier as there is no more room in the buffer.

The proposed BS architecture, on the other hand, gives much better performance as the intelligent classifier stops accepting packets after a threshold level is reached in the buffer, thus leaving a lot of room for the good packets.

4.2 Scenario 2 - Large Enough Buffer (Burst of 200 Banned packets):

In this scenario the base station receives a random order of packets until a point when it is attacked with a set of 200 Banned CID Real-Time packets, but this time it has a large enough Real-time buffer to hold them all. The objective was to compare queuing delay.

Plots, displaying the queuing delay and state of the Real-Time buffer, for both the architectures (in parallel) are shown below:

As can be seen from 4-c, the queuing delays in both architectures overlap up to a certain number of received packets, after which there is sharp increase in the queuing delay experienced by packets in the existing architecture. The cause of this behavior is evident from the plot below ( 4-d), which displays the real-time buffer state for both the architectures. The classifier in the existing architecture admits all the 200 packets, thus stuffing the buffer, and the following good packets experience a significant increase in delay. Comparing these results with the proposed architecture's, the delay stays consistent throughout, as the intelligent classifier stops accepting packets with Banned CIDs after the threshold is reached, thus not causing any significant queuing delay to good packets.

4.3 Scenario 3 - Large Enough Buffer (Regular data):

In this scenario the base station receives regular random data, with one banned CID and has a large enough Real-time buffer to hold them all. In this scenario, we experimented with the size of the packets by increasing the size from 6 bytes to 40 bytes. The objective was to compare queuing delay.

Plots, displaying the queuing delay and state of the Real-Time buffer, for both the architectures (in parallel) are shown below:

In 4-e, it can be noticed that the proposed architecture's intelligent classifier


[1] G. S. V. Radha Krishna Rao and G. Radhamani, WIMAX - A wireless technology revolution, NewYork : Auerbach Publications, 2008

[2] Sanida Omerovic,Faculty of Electrical Engineering, University of Ljubljana, Slovenia, “WiMAX Overview”,

[3] Aravind Seshagiri, Research Analyst, Semiconductor Group, “Inbuilt QoS - Advantage WiMax” April 23, 2005,

[4] Tarik Taleb, “Toward Efficient Service-Level QoS Provisioning in Large-Scale 802.11- Based Networks” IEEE Network, September-October 2007.

[5] WiMAX Tutorial,

[6] Westech Communications Inc., WiMAX Forum, “Can WiMAX Address

Your Applications?” October 24, 2005,

[7] Ed Grabianowski and Marshall Brain, “How WiMAX Works”

[8] Last mile,

[9] WiMAX,

[10] Mohammed Dawood, “WiMAX & QoS”, May 2007,

[11] What Every Company Needs To Know About Mobile WIMAX and QoS,

[12] Tom Evslin, “WiMAX vs. WiFi” Feb 20, 2008

[13] N.F. Mir, Computer and Communication Networks, Prentice Hall, Nov.26, 2006

[14] Eric Meyer

[24] A hierarchical mobility management framework for the Internet, Castelluccia, C.; Bellier, L.; Mobile Multimedia Communications, 1999. (MoMuC '99) 1999 IEEE International Workshop on, 15-17 Nov. 1999 Page(s):149 - 154

[25] Saving Public Addresses in Mobile IP, Zhen Zhen;Networking, International Conference on Systems and International Conference on Mobile Communications and Learning Technologies, 2006. ICN/ICONS/MCL 2006. International Conference on, 23-29 April 2006 Page(s):127.

[26]Web: Mobile IP: Enabling Mobility for 3G wireless, Paresh Jain and Rakesh Kelkar.

[27] RFC 3220, IP Mobility Support for IPv4.

[29] C. Perkins, Mobile IPv6 & Cellular Telephony

[30] Mobility Support in IPv6, draft-ietf-mobileip-ipv6-16

[31] What is Bluetooth?, McDermott-Wells, P.; Potentials, IEEE, Volume 23,  Issue 5,  Dec 2004-Jan 2005 Page(s):33 - 35

[32] Piconet: embedded mobile networking, Bennett, F.; Clarke, D.; Evans, J.B.; Hopper, A.; Jones, A.; Leask, D.; Personal Communications, IEEE [see also IEEE Wireless Communications], Volume 4,  Issue 5,  Oct. 1997 Page(s):8 - 15

[33] Implementation of novel application using Bluetooth, Kyuhwan Seol; Yeongseog Lim; Volume 2,  4-6 Aug. 2004 Page(s):1617 - 1620 vol. 2

[34] Interworking Wireless Mesh Networks: Performance Characterization and Perspectives, Tsai-Wei Wu; Hung-Yun Hsieh; Global Telecommunications Conference, 2007. GLOBECOM '07. IEEE, 26-30 Nov. 2007 Page(s):4846 - 4851

[35] A Light-Weight Client Mobility Approach for Infrastructure Mesh Networks Wishart, R.; Pirzada, A.A.; Portmann, M.;Networks, 2007. ICON 2007. 15th IEEE International Conference on 19-21 Nov. 2007 Page(s):252 - 258

[36] Wireless Mesh Networks by I.F Akyildiz web:

[37] A survey on wireless mesh networks, Akyildiz, I.F.; Xudong Wang;Communications Magazine, IEEE Volume 43,  Issue 9,  Sept. 2005 Page(s):S23 - S30

[38] GStokes: A Grid-enabled Solver for the 3D Stokes/Navier-Stokes System on Hybrid Meshes Langer, U.; ulehner, W.; Huidong Yang; Baumgartner, M.;Parallel and Distributed Computing, 2007. ISPDC '07. Sixth International Symposium on 5-8 July 2007 Page(s):50 - 50

[39] Multimedia-Centric Routing for Multiple Description Video in Wireless Mesh Networks'Shiwen Mao; Hou, Y.T.; Sherali, H.D.; Midkiff, S.F.;Network, IEEE Volume 22,  Issue 1,  Jan.-Feb. 2008 Page(s):19 - 24

[40] Stochastic Security in Wireless Mesh Networks via Saddle Routing Policy,Li, Xiang-Yang; Wu, Yanwei; Wang, WeiZhao;Wireless Algorithms, Systems and Applications, 2007. WASA 2007. International Conference on 1-3 Aug. 2007 Page(s):121 - 128

[41] Book: High-Speed Networks and Internets 2nd Edition by William Stallings.

[42] WF2Q: worst-case fair weighted fair queueing,Bennett, J.C.R.; Hui Zhang; INFOCOM '96. Fifteenth Annual Joint Conference of the IEEE Computer Societies. Networking the Next Generation. Proceedings IEEE,Volume 1,  24-28 March 1996 Page(s):120 - 128 vol.1

[43] - The Infrared Data Association Home Page - the main source of information about the IrDA including the actual specification..

[44] “Channel-Aware Scheduling for QOS and Fairness Provisioning in IEEE 802.16/WiMAX Broadband Wireless Access Systems” by Iera, A.; Molinaro, A.; Pizzi, S.; Calabria, R. Page(s): 34-41