QoS for 2.5G and 3G Networks
QoS for 2.5G and 3G Networks
Jan. 1, 2000 12:00 AM
Analyzing the customer experience to ensure that the end user has a perception of value creates an environment in which a carrier can maximize profitability and a return on investment.
Over the last three years, much progress has been made in the field of mobile interactive services, from building simple services such as ringtones and icons for enhanced handsets, to today's MMS over GPRS or CSD (circuit-switched data).
During this evolution, operators, manufacturers, and CSPs (content service providers) have looked for killer applications, business cases that let the VAS (Value Added Services) be as profitable - or more profitable - than voice call services. The financial world looked at these developments with interest. Analysts focused their attention on VAS potential, ignoring economic fundamentals that didn't have such an effect on stock sales. Now telephone company stock trends depend mainly on VAS EBITDA (earnings before interest, tax, depreciation, and amortization) variation.
In such an environment, telecom companies worked hard on developing interactive services. In Japan, NTT DoCoMo defined and deployed i-mode. In Europe, the standardization proposed by TIM (an Italian GSM operator with extensive worldwide operations) to the GSM Association gave birth to m-services, implemented by handset manufacturers and operators. M-services enabled some interesting features key to VAS business development such as EMS, WAP Push, and MMS.
NTT had much success with i-mode and thus started collaborating with European operators to launch i-mode services in Europe. Nevertheless, the greater part of VAS revenues in Europe are still represented by mobile messaging (actually SMS and MMS). Why? Because of the QoS perceived by the end user!
The value of messaging in the marketplace is shown by the large usage of SMS and MMS services: this means that the quality standard satisfies the customer both for usability and value perceived versus price - also for daily infoSMSs that cost the end user 55¢ each.
In messaging services, the end user receives an alert of the new content (i.e., the ringtone of a new message) after the download in the handset is completed. Then he or she can use the object downloaded immediately. No broadband is required, even if the content consists of a large amount of bytes. The only limiting factor is the delivery time delay: a hundred Kbytes of content needs more time to be delivered than a 170-character SMS. However, this factor does not reduce the service usability.
Public networks bandwidth, especially for mobile communications, is affected not only by the technology capability but also by other parameters, not least of which is the transmission load over the network. In packet networks - GPRS and UMTS packet data - the situation could be more complicated for retransmissions due to errors.
The two remarks mentioned above give service developers the following guidelines: real-time services generally have to be designed to work in a narrowband channel and messaging services can be designed to work in broadband. Though this may appear to be common sense, many CSP manufacturers and telecom operators are focusing on real-time broadband services such as video conferencing or video mobile phone calls (a different approach would be to dedicate a broadcast or multicast mobile channel), giving customers the expectation of making a video call over the cellular phone.
In a mass market that does not provide diffused video telephony, the consumer expectation is to have clear video calls, anytime, anywhere - a great challenge over a mobile public network, due to both the bandwidth availability and infrastructure costs. Who can state that mobile video calls are not a killer application? Maybe the financial world assessing the break-even performance of the video call service offered at the expected price. Maybe the customer for the expensive service offered, set by the operator to ensure a rapid return on investment. Maybe the customer for disappointment due to the availability of the service and the quality of the moving images - in other words, for the QoS perceived.
What is the QoS perceived by the customer? Unfortunately, the QoS, in the customer point of view, is not an equation of the network parameters required for delivering the ser-vice. QoS for a customer is to receive the service expected (i.e., the service he or she thinks it should be) together with a standard that reflects the amount paid for it, in the same way a diner in a restaurant expects to pay a rate proportional to the quality of food, service, and ambience offered.
So a hypothetical mobile voice service called "Find a Pharmacy," creates the customer expectation that he or she can call an IVR (Interactive Voice Response) system - or an operator - anytime, anywhere, to be told the nearest pharmacy to the calling place. For an advanced customer, it can create the expectation that the mobile network can automatically locate his or her position - but, most important of all, is that the customer expects that the pharmacy indicated is open!
The customer who calls the service on Sunday and discovers that the nearest pharmacy indicated is closed, has a low quality perception of the ser-vice and can also lose money and waste time following the service's recommendations. How large is the investment required to maintain the database of the shift of pharmacies over national territory? What can the final price be considering the financial obligations? Many "killer applications" were revealed as "killing applications," because they disappointed customers who, as a result, didn't use the service again. This produces "VAS skeptical customers," starting a negative viral marketing action.
Last year, together with Openwave's consulting staff, I made a service effectiveness analysis on TIM's WAP services. One of the most important observations was exactly the behavior described above: many customers who experienced errors during their first connections to WAP services became "Lost Opportunity Users" (i.e., users who did not return to the service within a specified observation period).
Manufacturers provide sophisticated tools to monitor QoS. Telecom operators have advanced Network Management Centers that monitor QoS, providing quality reports and real-time alarms for troubles. This is necessary and highly profitable, but does not help to give a feel for the Quality of Service perceived by the customer.
End-to-End VAS Monitoring System
The system I proposed is composed essentially of three parts:
The probes of the monitoring system are represented by:
SLAM The Service Level Agreement Management system is based on the tracking of a TTS (Trouble Ticket system). In practice, with third parties and internally, TIM defines the ser-vice-level parameters provided to the end user. Parameters such as delivery time, integrity, and relevance of contents are associated with a "performance" percentage. For example, a financial alert on a company stock rise, has to be delivered within 30 seconds in at least 98% of the cases (availability) from the stock variation and must report the right final value 100% of the time. Responsibility for the service level is held all along the service supply chain. If one of the probes detects a failure within the 30 seconds or on the stock value, it sends a Trouble Ticket, starting the TTS process, which involves the entire supply chain to fix the problem. The TT has to be "closed" within 48 hours and all the elements of the supply chain have agreed on an SLA that includes penalties.
Quality Business Report
The TIM monitoring system will be fully operating before next summer. I believe it represents one of the best available solutions to ensure the QoS in the customer's perspective, and to assure the operator more congruency between market focus groups and budgets.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week