Monitor IP telephony -- With Probe Tool?
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
The initial goal of the ex-job was to verify whether it is possible to use a probe to guarantee a certain quality and to improve the quality of an IP telephony service. The definition of a probe is somewhat vague, but generally you can say that a probe is non-intrusive, and therefore does not affect the network. It is a tool that normally listens to all messages that are transported over a network and records statistics. Almost any kind of network monitor could be called a probe. As the ex-job proceeded it became clear that there were no existing probe tool that worked for measuring IP telephony. Thus I started to look at how an IP based network is normally monitored and what you can do to guarantee QoS (Quality of Service) in an IP based network.
IP telephony is based on digitizing, and possibly compressing, audio information into packets of data, which are sent over the IP-based network and finally reconverted from data to audio at the receiving end. Even though the technique for IP telephony is well established, no major telephone company has implemented more than pilot tests. This is because today it is not possible to offer controlled high quality IP telephony. The standards are not yet firmly established and the tools to help operators provide high quality service either do not exist or are still being developed. Thus different vendors' products do not work together. During the time of working on this thesis, a lot has changed, new products are being introduced to the market and standards are being established. It is said that 1999 will be the year for IP telephony, the year when it will be taken seriously and really boom.
One of the motives for IP telephony is reduced cost. The equipment is cheaper and it gives users the chance to bypass the ordinary telephone network, which is sometimes expensive, for example for international phone calls. Compared to normal PSTN the cost is about 40% lower, even when looking only at the comparison with circuit costs and this does not include the cost savings for the much less expensive routers instead of multiplexers and exchanges that are used in the PSTN. Another motive, and probably the most important one is the flexibility of offering new services once you are able to send different types of data over the same network. When all traffic is packetized in IP packets also makes it easier to add new services to the edge of the network. The goal is that the callers will notice no difference between IP telephony and PSTN telephony.
Packet networks have a long way to go to deliver high-quality voice services, but this is not the only obstacle on the way before IP telephony will be fully developed. For customers to be interested in the service, some kind of added-value has to be offered. For the telcos (telecommunication companies) IP telephony was for a long time a dilemma; if they would invest a lot of money in IP telephony and launch a high quality product, they would actually be competing with themselves. But if they did not do anything, they would risk being left behind as the new technology evolved. Strategic and tactical moves will be made every step of the way. Hundreds of billions of dollars have been invested in the circuit switched technology and the companies that have made these investments will not be anxious to retire these investments before they have shown profits. It could take anywhere from five years up to more than a decade to effect such a massive changeout. In fact, we don't know if the transition from one network to the other will ever be complete. More billions of dollars will have to be invested in packet switched equipment, and test equipment has to be developed. However, there appear to be many new operators willing to invest in this new infrastructure, so even if the existing telecommunication operators do not make this change others will offer their services.
Existing operators have to face the fact that IP telephone minutes generated by carriers in September 1998 totaled 35 million, based on data supplied by service providers. This sum represents a six times increase from 6.3 million minutes generated in December 1997. Expectations vary, but within approximately 5 years one third of all phone calls will be over packet switched networks. Knowing this makes it even more important to be able to offer this service with sufficient quality.
Place, publisher, year, edition, pages
1999. , 89 p.
IdentifiersURN: urn:nbn:se:kth:diva-95338OAI: oai:DiVA.org:kth-95338DiVA: diva2:527846
Master of Science in Engineering - Electrical Engineering
1999-04-21, Biblioteket, Teleinformatics, Isafjordsgatan 22, Kista, 16:00 (Swedish)
Maguire Jr., Gerald Q., professorSuta, SandorBergström, Olle
Maguire Jr., Gerald Q., professor