Quality of Service within Wireless ATM - Essay Example

Quality of Service within Wireless ATM Introduction As the demand for personal mobile devices and services increase, the popularity of ATM (Asynchronous Transfer Mode) networks have also increased. ATMs are a from of adaptable broadband communication systems, and with the coming of wireless ATM would greatly increase the ease of accessibility of advanced applications and other electronic functions while on the go. Wireless ATN is seen to go further than wired ATM, in terms of fewer delays, less data corruption and loss. There are perceived problems though such as interference from radio frequencies and these and other details of ATM will be discussed further (Lessandro et al, n.d.). In today's modern society, the two major electronic systems that are patronized today are Broadband internet connections and mobile devices and technology. For the future of broadband connections, ATM or Asynchronous Transfer Mode is seen as the best method of transferring data for Broadband Integrated Services Digital Networks (B-ISDN) mainly due to tits ability to support many types of services other than voice and music data. For indoor data communications however, LANs or Local Area Networks are more preferred because of their almost seamless and faster data transfer speeds. It is in this sense that the combination of wireless and ATM technologies would greatly enhanced LAN environments. (Passas et al, n.d.). There quite a number of studies that explore specific coding/retransmission techniques in wireless communication systems (Nakao et al, 1996; Lambrette et al, 1997). All of these featured protocol/control schemes are tested in so-called one-on-one scenarios and there had not been cases where the performance or behaviour of wireless ATM system would perform in a data/resource sharing environment or scenario that involves multiple units. These studies may have shown the capabilities of specific applications of wireless ATMs but it cannot be a basis for the use of wireless ATMs in a networked scenario. There are cases that some of the studies were conducted that are irrelevant or unrelated to the study of ATMs such as GSM/DECT tests and channel behaviour (a channel with iid errors is often assumed) (Lessandro et al, n.d.). ATM Overview Asynchronous Transfer Mode or ATM for short is a group or network of 'fast packet switching systems' that are linked through point-to-point node connections. An ATM 'fast packet' typically consists of a total of 53 bytes in which 48 bytes is dedicated to data while 5 bytes is dedicated to control information. In order for data to be transferred securely form these connections; a point-to-point connection must exist between the 2 host systems. The level of service between the 2 host must be set-up initially before data exchange can be facilitated between them (Bhalla, n.d.). Wireless ATMs are being seriously considered to be the next framework for next generation wireless connections which will be able to support various advanced multimedia services with different levels of Quality of Service (QoS) requirements. There have also been thoughts on how wired broadband ATM systems can be extended into wireless ATM systems. This is mainly due to the fact that ATM provides the advantage of statistical multiplexing (flexible bandwidth allocation) and quality of service (QoS) that other systems cannot provide. This move is suggested because of the increasing demand for mobile devices like portable computers, cellular phones and other similar devices. Modern consumers rely on these devices for daily transactions and other activities so the rate of development for incorporating ATM systems is considered as the next big leap not only for communications but also for the future of human society itself (Osama et al, 1997). Wireless ATM Wireless ATM is similar to wired ATM with the addition of several new features such as increased mobility for the user, dependability and the ability to support various media type effortlessly. Combined with micronized cellular technology (ATM cells size range from the micro to pico levels), wireless ATM would sure provide faster data transfer without much delay or interruption Mitts et al, 1996). The network's mobility aspect comes from the 'decoupling' or separation of the normal mapping scheme for the node and switch port. The set of wireless nodes connects to a wireless access point which is served by a single ATM wireless switch (Awater and Kruys, 1996; Mitts et al, 1996; Bhalla, n.d.). There are two approaches in which wireless ATMs can be integrated to complement a wired ATM network. The first approach is that a wireless ATM can be used as an 'overlay' in which support for mobile systems is usually done using independent network elements that adds to the mobility of the device. The alternative approach is that the wireless ATM is integrated in the network itself. In this approach, the ATM is equipped with mobile specific features. This results to a network that can cater to both mobile and fixed systems, in addition to fact that there are no special switches required to support mobility. The fixed ATM network then becomes a switching and transmission infrastructure that caters to both fixed and mobile clients. It works by having the access point connected to a switch with an ATM UNI interface which can be shared between any number and combination of fixed and mobile users. Location management, terminal authentication and other mobile functions are also located in the switch. These are essential for the mobile terminal while another essential protocol called the Access Point Control Protocol (APCP) is used to allow the switch to interact with the connection set-up and hand-over applications (Bhalla, n.d.). The end-user(s) with a Wireless ATM Mobile Terminal can now access the Mobile ATM network through Wireless Access Points (WAP) which then allows the user to communicate with other wireless devices and networks which are connected to other ATM switches (Jiang, n.d.) Both synchronous channel networking (PDH, SDH) and packet-based networking (IP, Frame relay etc.) is supported by wireless ATM, while also supporting multiple QoS for the flow of data packets. This way, circuit-switched and packet-switched networks traffic is resolved by the mapping of both bitstream and packet streams into a stream of units called 'cells' that are marked or 'tagged' with virtual circuit identifiers. These 'cells' are sent by demand to the system within a set time pattern in a parallel 'bit-stream'. What happens is that there is an asynchronous sending of the cells rather than the (low-level) bitstream that carries the cells (Wikipedia, 2006). Originally, ATM was intended to be the technology behind Broadband Integrated Services Digital Network (B-ISDN) that would replace the existing PSTN. A full complement of ATM standards provides characterization for layer 1 (physical connections), layer 2 (data link layer) and layer 3 (network) instead of the classical OSI seven-layer networking model. In a way, ATM standards drew their ideas from telecommunications instead of computer systems because computers are more likely to communicate with each other rather than being just connected. It is in this light that the integration of telecom characteristics were adapted instead of normal computer system analogies. ATM ha s thus been a success thus far in their applications as a medium of transport for IP traffic for private LANs and international telecom networks but it failed in its goal of being an integrated technology for LANs, public networks and other user services (Wikipedia, 2006). Wireless ATMs can run ATM applications if there are specified and proper protocol layers that provide uninterrupted information exchange through radio waves without sacrificing QoS while being compatible with the right ATM. This way, wireless ATM aims to 'fake' the operation of applications by providing almost the same interface of a fixed ATM to the AAL or ATM adaptation layer. As mentioned earlier, application messages are first divided into TCP'IP packets and then into cells in which the wireless ATM simply converts these ATM cells into a format that can be transmitted via radio to transport it to the other end of the mobile terminal link. Upon the arrival of data on the other end, he original ATM cells are reassembled and forwarded into the nodes of the ATM circuit through the network connection (Lessandro et al, n.d.). Connection behaviour considerations There are three main effects of a wireless ATM connection to mobile users. The first effect is that there is a significant difference in QoS between wired and wireless ATMs. There is almost no interruption of QoS in a wireless connection than in a wired connection. The second effect involves rerouting of connections due to the aforementioned handovers. The effect is not really about connections but rather about the coverage of mobile specific information about the kind of connection that is required to be established at that particular location (Mitts, 1996). Connection renegotiation usually occurs and is primarily a concern for the user or to be more exact, for the applications that the user currently runs. The user can set the application to optimise its performance, allowing it to counter any notable QoS changes as much as possible. The reaction of the application depends on the data content, and changing the bandwidth or data content may make data traffic better in this case. Other options of improving data exchange rate are to simply lower the data rate as done in video data exchanges (Apteker et al, 1995). Lowering the bit or frame rate depends upon the situation and can either be used as a way to increase the coding used for error connection (FEC) and bandwidth at the same time or it could be used just to lower the bandwidth requirement for the connection. It is equally important to specify the type of problem and at the same time the level of change to establish a proper connection for QoS renegotiation and complete control over data traffic. (Mitts, 1996). Connection modification, in theory, could also be done and be quite useful in fixed ATM networks. If this can be possible, then mobile connection requirements will decrease. This is predicted to be initiated by the users, with the mobile network being the initiating party that would control resource sharing (Mitts, 1996). On the other hand, handovers create a whole lot of different issues. On the user's side, a handover is intangible and is just viewed approximately not unlike for an application which considers it as very tangible. If this is the case then a handover then becomes a matter of network implementation rather than something else (Mitts, 1996). There are possibilities and situations in which the user and application wants to control the performance of handovers. An example of this is the handover between different operators which results to different connection charges because of the handover. A user which uses an application or a set of applications would preferably want to minimize or even inhibit handovers to minimize charges. This is often true when the handover is from a private network which is eventually distributed over a public network (Mitts, 1996). The final factor involved in wireless ATM connections is radio frequency issues. These issues commonly includes information about the possibility of initiating a connection, QoS information especially if the network is running multiple applications, the user's current location especially if the service is location dependent or location sensitive (Mitts, 1996). Radio Channel Characteristics In the modern s scenario, mobile ATM links are still unreliable due to poor data trafficking abilities not unlike fixed connections (e.g. optic fibres). There are several solutions being formulated for mobile radio interference, and these errors are attributed to the following: (Skalr, 1997) path loss, interference, shadowing and fading. The negative effects of path loss, interference and shadowing can be alleviated by proper cell planning to make sure that the carrier-to-interface ratio (CIR) is tolerable in the given area of operations. This can be done by examining location data and other methods such as 'ray tracing', a method that can accurately predict power distribution, optimize cell size as well as the base station position. Situations such as heavy 'shadowing' by mobile objects can now be modelled and planned which can lead to proper signal detection even with poor CIR (Lessandro et al, n.d.). Fading model There are two types of environments in a fading scenario: multi-path fading environment and flat fading environment. In a multi-path environment, each signal is received in multiple delayed batches which differ in phase and frequency (due to the Doppler Effect). In a flat fading environment (as in a broadband connection), the relative location of the device and the obstacles are what affects transmission speed (Lessandro et al, n.d.). Transmission model Studies show that wireless channels can exhibit wild behaviour depending on the network environment and conditions such as cell size, geographic features, available power, antenna type, terminal mobility, spectrum window (Bultitude and Leslie, 1993; Hashemi, 1993). Thus it is difficult to make a transmission model at the bit/symbol level with out the selection of a specific wireless ATM implementation. From this point, the general channel performance will be best described in terms of cell errors, modelled by a twin state Markovian chain. By calculating the factors in a chi-square test, the process can be proved by DECT frequency measurements (Lessandro et al, n.d; Zorzi and Rao, 1997; Roda, 1996). Security issues With the emergence of internetworking (connection between networks) technology, security in networking has been more of an issue than ever before. Internetworking may have its benefits but due to its open nature, it is very susceptible to various forms of digital attacks. Most network technologies of today such as ATM are inadequately equipped to deal with such attacks and one way to remedy this is by redesigning its entire security system (Liang, n.d.). ATM security system was never an issue and was never given any particular attention up until 1995 when an ATM forum group decided to clear up these security issues. Compared with other types of systems and their security measures it can be said that ATM security is still in tits infancy and a lot of other issues must still be addressed and cleared in order to have a working idea on how the whole security system for ATMs will work. Thus far, the threats to ATM networks and the requirements for their security have been discussed. The actual implementation of ATM security issues will now be discussed in the further (Hanson, 1995; Chuang, 1996; Deng, 1995, Laurent et al, 1997,) As an overview, the factors that threaten ATM networks are: eavesdropping, spoofing, service denial, VC stealing and traffic analysis etc. with VC stealing and traffic analysis happening specifically to ATM systems only Liang, n.d.). Threats to ATM networks As mentioned earlier the threats to ATM networks are the following: eavesdropping, spoofing, service denial, VC stealing, traffic analysis and many more. Note that VC stealing and traffic analysis are two threats that occur specifically in ATM networks. (Liang, n.d.). Eavesdropping Eavesdropping is a classic method of stealing information and it can be done simply by listening secretly to a conversation, tapping in a phone line or its more advanced equivalent, tapping into a computer network. Eavesdropping in a network happens when the attacker (or a 'hacker' in the modern colloquial) connects or 'taps' in to the transmission stream in order to gain unauthorized access to classified data. Eavesdropping is the most common form of attack that happens in an ATM network. There is a general belief that ATM networks are not easily tapped into since these networks are connected via optic cables. There are reports however that the average cost of tapping equipment for fibre optic cable networks can be acquired for just about $2000, which can be easily afforded by anyone. What is worse is that specialized knowledge is not a necessity when tapping unto networks, and most of this knowledge is widely available to hackers and any individual through academic and non-academic sources (such as the internet) (Bacon, 1989). For example, any member of an ATM forum can acquire access to the contributions and specifications about the architecture, structure and security of the ATM network. Knowledge on how to hack into an ATM network can also be acquired easily through online sources. There are also extrapolations that as technology advances, the level of spying unto other networks will also advance and security systems must be able to cope up with this advancement so that documents and other data filing types would not fall into the wrong hands. (Liang, n.d.). Spoofing A spoofing attack happens when the attacker or hacker impersonates or fakes the identity of another user to gain access or destroy another user's essential stash of data. This happens much in real life as it does in computer networks and is known to be one of the most devious threat forms ever devised. Unlike eavesdropping, spoofing requires more sophisticated equipment to manage the data as well as special access codes. Since most networks are connected through dubious internet connections, it is easy for hackers nowadays to access restricted data even more. ATM is no exception, since it is also connected to a public domain; it is very susceptible to this kind of digital attack (Liang, n.d.). Service Denial ATM basically depends much on the user's network connection, most commonly known as a Virtual Circuit (VC). A set of 'setup' signals manages an ATM which can be easily disconnected from the whole network by a 'release' or 'drop party' signal. These two types of signals tell any nearby switch within the ATM thereby disconnecting the VC (Stevenson et al, 1995). Hackers can send these signals as often as they want to the system, providing interference and disturbance to the normal flow of data transfer between users. This method can be used in combination with any other threat (such as eavesdropping and spoofing) and can greatly disturb or even completely block out a user from the services (Liang, n.d.). Stealing of VCs It is possible for a hacker to steal a VC if two switches in an ATM compromise each other. Assuming that there are two VCs, VC1 and VC2 which will go through a pair of switches; Switch A and Switch B. Assume also that VC1 is used by User1 and VC2 is used by User2. if Switch A and B compromise each other then switch A can switch VC1's cells going from A to B through VC2 and switch B will switch back those cells to VC1. Switches move cells based on their VCI (Virtual Channel Identifier) or VPI (Virtual Path Identifier) in the cell header, so switch A and switch B can just alter these fields to and from each other. These changes are not able to be detected by switch A and switch B and will switch the assumed VC2's cells just like the original VC2's cells. In public networks that switches packets often; User 1 will technically benefit a lot from this manipulation by 'stealing 'the higher quality channel assigned to both users. Also user 1 (assumed to be the hacker) can take a great advantage of user2 if user2 is the paying client for the connection. Some experts believe that it not entirely impossible for the two channels to compromise each other. But in any case, this situation is only possible if the network is owned by just one company, but in an internetworking environment, the rate of compromise is high especially if the case cells travel at different ATM networks from time to time (Liang, n.d.). Traffic Analysis Hackers can also acquire restricted data by collecting and analyzing the information like the volume, timing and the communication patterns of a specific VC. Volume and timing can disclose a wealth of information to the hacker even though the data is encoded because the encoding does not affect the volume and timing of the data. The hacker can also trace the source and destination of the data by obtaining and analysing the cell header which is written usually in plain text as well as using some knowledge about the routing table. In relation to this, covert channels can also be used, in which the hacker can encrypt the information of the timing and volume of data. , VCI or session key to discharge unnecessary information to a user without being traced or monitored. Normally, in a loose network, this kind of threat will not happen but it commonly happens in a strict security environment (Liang, n.d.). Establishing a security system is a good way of preventing these attacks and counts as a good management measure. One of the security measures that can be done is the encryption and decryption of data. Using encryption/decryption measures enables a more secure connection and prevents the unauthorised access of data by hackers. Aside from decryption/encryption, a network system should have automated or semi-automated key management and distribution as key distribution will eventually go through the network. The only issues thus far that can emerge from this are how to secure the keys when these move over the network (Liang, n.d.). QoS considerations The main concept behind an ATM network is to provide necessary connections with a sensible Quality of Service (QoS). The main difference between a wired and wireless ATM system is that wired networks require a more physical connection between users. Wireless ATM networks on the other hand offer the user mobility because of lesser physical constraints involved. With this characteristic, the QoS for a wireless network is hard to determine, so it is theoretically possible for wired ATM concepts to be applied to wireless ATM connections in the future (Mitts, 1996). Specifying the Quality of Service in the current ATM application environment present some major problems and benefits. One of the major problems of current QoS for current fixed networks when applied in a wired network is that the connection is decisively established at the initial time of connection and this is expected all through out the time that the connection runs. With this, maintenance of the QoS of the wireless connection remains to be a challenge in today's standard (Mitts, 1996). Interpretation of a QoS objective in a wireless system. The very first requirement for the application of ATM QoS concepts to wireless ATM connections is to create and define a more precise definition of a QoS objective or guarantee to be applied in a wireless network. A QoS guarantee of about 99.5% most of the time is still considered as a 'statistic' only and that this kind of guarantee is quite possible and totally acceptable for the user. However a 15% QoS standard guarantee is obviously not acceptable to a user and can naturally bring about negative remarks from the user (Mitts, 1996). The traffic contract in a fixed ATM network is the prime specification for the QoS that the network must provide to the user and is therefore is a responsibility of the network operator/ administrator. This works by the assessment by the network of the connection's acceptability during the initial call acceptance phase. A call is acknowledged or discarded based on these estimates. The predictable nature of wired networks makes this assessment possible and acceptable. For fixed ATM network, the various mechanisms that affect the overall QoS of a connection are switches and transmission links which do not have the same dynamic behaviour except for the difference brought upon by the statistical nature of the traffic and ATM functions. These differences should be measured and managed by statistical analysis (Mitts, 1996). In a wireless (mobile) network, the situation gets more complicated. There are many characteristics that make it to maintain QoS in wireless networks, such as physical limits that make the assessment of QoS nearly impossible or non-applicable. First of all radio interference introduces considerable difference in the level of attainable QoS due to fading and shadowing, range limit of the radio connection and the mobility of the user subject to different environmental factors. A user that utilises the mobility of the network moves between cells. In this situation, cellular systems use different cells that provide different QoS at maximum levels which is because of different load speeds and other factors. At anytime a call is accepted, it is nearly impossible to determine what radio cells the user will use during the time of call and what cells can be provided the cell can provide when a user eventually enters it. Thus the initial traffic contract and specifically some structures of QoS level of agreement, which is more or less not of any worth to the user after the user hands over to another cell. A sort of reservation of resources can be done to solve the problem partly but at the cost of less efficient resource use and increased complication of the network system (Acampora and Nagshimeh, 1994; (Mitts, 1996). Alternative interpretation It is thus very essential that a significantly different approach or definition of the Quality of Service (QoS) is to be implemented for mobile networks as compared to fixed networks. The following paragraphs will discuss the various possible alternatives for the definition of QoS guarantees in a mobile scenario (Mitts, 1996). The different approaches that can be applied to mobile networks are as follows: The first approach is the 'tough luck approach'. There are no explicit QoS guarantees that are available in the current scenario for mobile networks. So in this situation, the traffic descriptor would only be defined by the network as a gauge of how the resources would be needed and utilised as well as how it would be beneficial to the operator. The user would then need to specify the QoS because the QoS is not guaranteed in the first place, especially in fixed situations. This can be utilised for the introduction of a new QoS class for mobile connections providing a sort of 'what you see is what you get' solution (Mitts, 1996). The second is the limited liability approach. In this approach the original QoS requirements of a given service are done but with some special conditions, such as when the user moves to the different INS and out of a cell or transferring from one type of cell to another (which is usually worse than the average QoS that the previous cell). This situation can be classified under the special mobile subclass of QoS which say that due to physical constraints of radio frequencies, the quality of Service experienced from that can be worse than normal. In this case, the user does not get negative QoS feedback while the applications that the user uses would not automatically react to QoS degradations. However, this approach requires a reactive and very flexible radio interface (Mitts, 1996). The third approach is renegotiated QoS. In a wireless ATM network could be either downgraded or upgraded (in other words variable) the QoS to equalize the rate of performance that is to be expected of the QoS for a mobile network. A renegotiation is frequently done after or in an ongoing handover process. This approach is done due to the belief that this approach will enable applications involved in the communication to react and solve transmission problems. Receiving a lower QoS could enable a video encoder to place more redundant data to the stream or even reduce the frame rate on the fly as an example (The ATM Forum, 1996; Mitts, 1996). The definition of a QoS is not just to be considered from the satisfaction or system efficiency view from the user. The application of QoS-based service could have some beneficial effects on the operation but this could be potentially the worse problem that can be encountered when used in public networks. If the requested QoS can be maintained and used then the user can be justified of the rates that they are paying. However, if a flexible mobile radio interface to be used by the user and just how much the abilities of this system can be used to provide a user with the QoS that are agreed upon which that can function even in the most adverse conditions, then it could be a basis for charge rates in a wireless system that can be potentially less than the charges for wired networks (Mitts, 1996). Conclusions The recent trend for all communication and computing devices today is mobility and compactness. Mobile phones have illustrated this by providing mobile subscribers with the convenience and benefits of compact and mobile communications systems. Computer systems are currently going to this path by introducing mobile technologies such as wi-fi (wireless fidelity) and wireless Asynchronous Transfer Mode (ATM). The concept of computing anytime, anywhere without interference is an objective that researches have been aiming for. Other benefits include lesser operations costs as well as convenience. Although this technology is now available and feasible today, it is not with out its disadvantages. One of these is physical constraints involved in the transmission of data through radio signals. Terrain features such as buildings and other structures are one of these constraints and the utilization of wireless ATM will make it possible to beat these constraints. Also, threats concerning wireless ATM network systems must be taken seriously. Threats such as eavesdropping and spooking are lesser threats which can be easily countered but more advanced treats such as VC stealing are hard to detect and counter so countermeasures should be formulated so that the proliferation of these unscrupulous would be lessened or eliminated thoroughly. The future of wireless computing is looked upon as the next big thing in computing and is looked forward upon by professionals and common people alike. If this happens then there are a lot of possibilities and potentials that will open up with the advent of these advanced mobile technologies from mundane activities to the more advanced activities such as researches and other scientific endeavours. References Acampora A. and M. Nagshineh. 1994. An Architecture and Methodology for Mobile-Executed Handoff on Cellular ATM Networks. IEEE Journal on Selected Areas in Communication. Vol. 12. No. 8. Apteker, R. et al. 1995. Video acceptability and frame rate. IEEE Multimedia. Vol. 2. No. 3. Awater, Greet A. and Jan Kruys.Wireless ATM - an overview . ACM/Baltzer Journal on Mobile Networks and Applications (MONET), Vol 1, 1996. Bacon, M., Security: a question of confidence, Telecommunications (int. ed.) (USA) Vol. 23, No. 11, pp 51-52, Nov. 1989 Bhalla, Reshma (n.d.) Wireless ATM - Routing and real time delivery. Retrieved May 10 2006, from http://crystal.uta.edu/kumar/cse6392/termpapers/Reshma_paper.pdf Bultitude, R. J. C. and A. W. Leslie. Measurement-based probability of error predictions for digital land mobile radio channels. In International Conference on Universal Personal Communicatons, pages 600-609, 1993. Chuang, Shaw-Cheng: "Securing {ATM} Networks",3rd {ACM} Conference on Computer and Communications Security, New Delhi, India, 1996, pp.19-30 Discussing challenges, secutiy mechanism placement and security requirements of ATM networks. Deng, R. et al: "Securing Data Transfer in Asynchronous Transfer Mode Networks"; Proceedings of GLOBECOM'95, Singapore, November 13-17, 1995, pp. 1198-1202 Dicussing security requirements and security architecture for ATM networks. Hanson, L., "The Impact of ATM on Security in Data Network", Proc. of Compsec International 1995, Conf. 12, pp 318-324 Hashemi, Homayoun. The indoor radio propagation channel. Proceedings of the IEEE, Vol. 81(No. 7):943-968, July1993. Jiang, Fan. (n.d.). A framework for using Simulations in Wireless ATM handover solutions. University of Jyvskyl. Retrieved May 10 2006 from www.cc.jyu.fi/anmima/kurssit/ semma99/FanJiang.doc Lambrette, Uwe, Lars Brhl, and Heinrich Meyr. ARQ protocol performance for a Wireless high data rate link. In Vehicular Technology Conference, pages 1538-1542, 1997. Laurent, Maryline, Olivier Paul, Pierre Rolin," Securing communications over ATM networks", IFIPSEC'97, Copenhagen, Denmark, May 1997 LESSANDRO,A., C. ANNARSI, M. ARTINO, DE MARCO, A. CHILLE, P. ATTAVINA. (n.d.). Quality of Service Issues in Extending ATM to Wireless Networks. Retrieved May 10, 2006 from http://www.aei.it/ita/ett_final/ett_word_letter.pdf Liang, Donglin. (n.d.). A Survey on ATM Security. Retrieved May 10 2006 from http://www.cse.wustl.edu/jain/cis788-97/ftp/atm_security/ MacKie-Mason, J and H. Varian. 1995. Pricing congestible network resources. IEEE Journal on Selected Areas in Communication. Vol. 13. No. 7. Mitts, Hkan (1996) Architecture of wireless ATM. Retrieved May 10 2006, from http://lib.tkk.fi/HUTpubl/publications/mitts96.html#C17 Mitts, H., H. Hansen, J. Immonen, and S. Veikkolainen. Lossless handover for wireless ATM . Mobile Networks and Applications (MONET), Vol. 1, No. 3, 1996. Nakao, Takayuki, Miki Yamamoto, Hiromi Okada, and Hiromasa Ikeda. Performance evaluation for error control scheme in Wireless ATM networks. In Technical Report of IEICE, pages 29-34, Institute of Electronics Information and Communication Engineers, 1996. Osama, Kubbar & Hussein T.Mouftah. (1997) Multiple Access control protocol for wireless ATM: problems definition and design objectives. IEEE Communications Magazine. Vol: 35 - (Issue11) Passas, Nikos, Sarantis Paskalis, Dimitra Vali, and Lazaros Merakos. (n.d.). Quality-of-Service Oriented Medium Access Control for Wireless ATM Networks. University of Athens. Retrieved May 10 2006 from http://cgi.di.uoa.gr/passas/IEEE_Comm_paper.pdf. Roda, Davide. Meccanismi di gestione delle risorse in retiradiomobili cellulari basate su tecnologia ATM. Ms degree thesis, Politecnico di Milano, 1996. Sklar, Bernard. Rayleigh fading channels in mobile digital communication systems, part I: Characterization. IEEE Commun. Magazine, Vol. 35, No. 7, pages 90-100, July 1997. Stevenson, D. and N. Hillery and G. Byrd, Secure communications in {ATM} networks Communications of the ACM, Volume 38, No 2, pp 45--52, Feb, 1995 Discussing issues like security threats, cell encryption and securing call setup in ATM network. The ATM Forum. 1996. ATM traffic management specification 4.0. Wikipedia. (2006).Asynchronous Transfer Mode. Retrieved May 10, 2006 from http://en.wikipedia.org/wiki/Asynchronous_Transfer_Mode Zorzi, Michele and Ramesh R. Rao. On the statistics of block errors in bursty channels. IEEE Transactions on Communi- cations, Vol. 45(No. 6):660-667, June 1997. Read More