If there is anything really new in IP telephony this winter, it is a cold sense of reality. A realisation that there is no real 'killer app', that migration to IP will take time, and that billing and cost issues need to be well thought out in the mind of the customer before the new era begins in earnest. Despite these problems, IP telephony is evolving and the incumbent telcos, new entrants and data networking vendors alike are jostling for key positions in the market. Chris Wyles, vice president and managing director EMEA for IP telephony vendor Clarent Technologies, said unified messaging, with a range of value added IP services, is being developed and deployed. He added that with clearing houses already established, such as Telia and AT&T - which allow operators to trade traffic and terminate calls where they may not have a gateway - IP telephony could easily become global. "The new operators are choosing IP because of its significant bandwidth capabilities and its low cost of entry compared with switched systems," said Wyles. "Gateways are getting larger and denser, with more ports in a given amount of space. Applications are being built that revolutionise network management, back-end systems and enhanced services." According to Wyles, the new devices will be able to handle phone traffic mixed with data traffic, such as cable modems, DSL modems and wireless devices, opening up a myriad of possibilities in terms of the structure of the IP network. Support for VoIP According to John Holdsworth, product marketing director at networking integrator Protek, Voice over IP (VoIP) hasn't yet found wide acceptance, although many traditional PABX vendors, such as Mitel, have been working to move the core switching intelligence to servers. Holdsworth said: "Integrated call distribution technologies are well suited to making use of VoIP now, so that both calls and data can be intelligently moved around the call agents, as well as being integrated with Web technologies to give more than one means for customers to obtain information in the optimum manner." Another area of focus is the delivery of Quality of Service (QoS) in the Wan to support services. But while there is a lot of discussion in the market about how to do it, there are few installations in existence. Holdsworth added: "The use of native mode IP networks to support these services does not yet seem to be a feature of the market. Companies using VoIP internally are in fact buying 'wires out of the wall' to carry their service mix, and then managing the IP flows using packet shaping devices to prioritise traffic appropriately." With the release of Windows 2000 and its promised QoS support, expected in the first quarter of 2000, there will undoubtedly be an upsurge in the use of IP bandwidth management technology and products that help manage the flow of voice and data traffic. A lot of the prerequisites are already in place, said Mark Randall, managing director of IP services distributor ATL Networks. So is the evidence that IP is more cost-effective than traditional technology. "IP services company Level 3 has calculated that to send a CD-ROM's worth of information from New York to Los Angeles costs $1.98 via an ISP and $27.08 via PSTN," said Randall. "While voice and data now account for about the same amount of bandwidth utilisation, the total cost of carrying all global voice traffic is estimated by IDC at $527bn per annum, compared with only $55bn for data." Furthermore, IP price/performance is improving rapidly, said Randall. This is because it is open and market-driven, not closed and proprietary like traditional telephony. "Doubling performance/cost takes 80 months for circuit switched technology compared with only 10 months for IP," said Randall. With competition increasing and consolidation forming crossover between merged vendors such as Nortel and Bay, R&D investment is likely to focus on a single platform capable of carrying both voice and data. "Integrated voice, video and data applications will provide real commercial benefits to integration of the two disparate environments. CTI, 'call me' and unified messaging applications are the earliest examples, but many more will begin to emerge," said Randall. The fact that IP telephony is available and in use today is not yet recognised by many people, he added. "An estimated 200 million IP telephony minutes will be carried in 1999 by Clarent hardware alone. According to some estimates, 2.25 per cent of US IDD traffic was carried on IP networks in 1998 and this will increase to 4.6 per cent in 1999. These are small numbers, but the growth is rapid and the fastest adoption is currently within telcos with the highest service level requirements." In most cases, deployments are on private IP networks or a single tier-one ISP. However, QoS issues are being addressed through the development of traffic management protocols such as DiffServ, and the remaining issues of reliability and functionality will inevitably be removed with time. This does not mean PBX and PSTN technology will disappear overnight. "The 'last mile' issue will not be resolved until IP is brought into the home via ADSL, wireless or satellite technology," said Randall. "PBX investments typically depreciate over five to 10 years." Randall also thinks we will see an encroachment of IP in the voice market and a concentration of investment in IP as a unifying technology. "Investment will initially be targeted towards PSTN/IP gateways and similar equipment that leverages existing investments. But as we approach 2000, it is clear that traditional telephony has no role to play in the future of the net century." Phil Ward, technical and support manager for telco Global One's Metropolitan Services Division, said: "The move to a fully integrated network will not happen overnight. The building blocks are in place to begin this task, but legacy networks have to be considered. Organisations still running X.25 and SNA applications are in the process of migrating to the latest protocol, which is IP in many cases. When vendor equipment becomes fully interoperable, networks will be able to benefit from multiple vendor components, which allow for local nuances." Migration has to be gradual, said Wyles. "The first step, which is happening quickly and globally now, is to put IP telephony in the service providers' backbone and into wholesale networks. It will take longer to bring converged technology to the local loop, because end users want to use their traditional telephones. Bringing IP to the local loop usually involves placing a new device on the customer premises." Wyles warns against assuming that traditional telcos are not internet savvy'. Traditional carriers and new telcos are finding ways to implement internet-related services, including IP telephony. Companies are more interested in providers that are in tune with developments, as the internet is now a vital component of successful business. "A company needs a partner that can understand the potential, offer 24x7 support and grow to meet needs," Wyles said. "However, some small ISPs do not have the resources and some of the bigger carriers are gearing up to offer comprehensive services. Many new carriers are faster on their feet and have the means to invest in the best technology." A recent report by IT consultant Ovum, entitled About Next Generation IP Networks: Service Opportunities from New Platforms, backed up Wyles' assertion. Iain Stevenson, principal analyst at Ovum, said incumbent carriers can use gateways to provide internet services, but cannot afford to wait before making a complete switch to IP. He predicted that within the next two years, new entrant telcos will be able to build a network using IP and gateway equipment alone. Howard Inns, marketing manager of network services and integration company Milgo, said: "The telcos will take some time to IP-enable their networks, and those in the vanguard will have the onerous job of convincing corporate UK of the benefits. So far, it's the SMEs that are blazing the VoIP trail." Integration in the wide area is happening, said Inns, but only on the back of cost savings, and it originates in the old 'telecoms mindset' which regards the network as little more than a costly overhead. "The message to these dinosaurs is that VoIP is not just about saving money," he said. "The network will deliver e-business benefits. Telecoms and IT directors will have to wake up to this when their colleagues in marketing and business development demand the extra business edge they can get from convergence." Looking for a killer When exactly the switch will happen is unclear. Although there has been much talk about unified messaging, Stevenson pointed out that there are currently no 'killer applications' for the integrated IP and PSTN network. He does not expect any application to catapult IP to dominance in any market segment. "There has not been a single user-driven application scenario that has compelled carriers to justify a significant investment. No application has proved capable of this so far and next-generation networks are no different." Even so, he believes the unification of services and the simplification of carrier infrastructure will help a range of applications open up the market for products. Established telco vendors such as Nortel, Lucent, Siemens, Alcatel and Ericsson need to enhance current Class 5 switches with packet add-ons, while adding integrated gateway products for new entrant carriers. Meanwhile, data network vendors with their heritage in IP, such as Cisco, face the challenge of developing products for incumbent carriers. Hence the trend for mergers and acquisitions which we must expect to continue for some time. IP BILLING - WHAT DOES THE FUTURE HOLD? Kevin Allodi, vice president of billing systems developer Portal Software, said there is no question that spending is shifting towards data services. "We expect the 'take rate' of enterprise IP voice services to increase dramatically in the coming 12 to 18 months. This will in turn drive further spending on IP/data infrastructure and support systems, to manage users and usage activity as well as to monitor for charge back purposes." Martin Browne, consultancy director at interconnect billing specialist Intec Systems, says that as 'free' internet access through ISPs and cable services becomes more widespread, operators will have to recover the costs of providing internet traffic - including VoIP - from a bundled monthly service charge. In the UK, IP traffic is expected to grow by 30 per cent per year, said Browne, while voice traffic is growing at about 5 per cent. It is expected that, eventually, voice will only account for between 5 per cent and 10 per cent of all network traffic. Voice is where operators make most of their money. "Even without significant VoIP traffic, transporting data can take up more than half the bandwidth but generate less than 12 per cent of revenue," said Browne. "The cost of providing stable bidirectional bandwidth to support VoIP will be prohibitive, but the revenue basis will be same as for any other data event." Once VoIP becomes widespread, traditional operators using switched networks may be forced to move from distance-based charging to flat rates. Once rating by destination goes, rating by duration and time of day could also disappear. Rating by duration makes sense for a voice call over a switched network, since the phone call is tying up a circuit for the duration of the call, but rating data calls by duration is less meaningful. The duration of a data call transferred over a packet switched network is more a function of the capability of the network and the speed of the user's modem. Pricing based on destination, time of day and, possibly, duration could be replaced by a combination of flat rates and rates based on the value and quality of the service delivered, be it VoIP or not.
Japanese researchers develop a flexible screen worn on the skin that they claim can monitor patients' heart rate and other vitals
ZenFone 5 Pro appears to boast a Snapdragon 845 SOC, an Adreno 630 GPU and 6GB of RAM
Pilot project will serve 300 homes to start with
The IoT faces significant compatibility challenges, which could be avoided for blockchain by adopting Hyperledger