|
|
article page | 1 | 2 | 3 | 4 | 5 |
…and in with the new
So today, the innovative but well proven strategies are:
$ Moving stable, unchanging applications already implemented in J2EE to open source servers
$ Wrapper-ing entrenched and difficult to replace applications with enterprise-managed web services
$ Building the applications which will provide core business functionality and competitive differentiation on pure SOA (and yes, you do need to invite those telecom managers to the design sessions to really understand the business functionality…)
$ Linking to users, customers, and supply chain partners via standardized web services.
Trends to be designing for right now
Several on the horizon technologies should be the new product development efforts in 2007. These are working but getting them into the mainstream will require the most sophisticated and advanced OSS/BSS that can be developed:
1: Deep packet inspection and Applications/content level routing
2: Single logical user identity
3: Viral VPN or social communities
1: Deep packet inspection began as a way of counting and forking internet traffic riding on high speed transports. Deep packet inspection is reading past the standard routing header of an internet packet and inspecting the application and port information usually used at the destination of packet transit. Then, decisions can be made on treatment of the packet. This is similar to how firewalls work, filtering based on deep inspection. If you can do this fast enough, you can use the information to differentially place packets in QoS queues or on VPNs. Devices are commercially available today to do this at all standard access speeds (but not at the highest speeds used in backbones). So today, you can purchase a customer access gateway which can prioritize all your VoIP traffic, your video conferences, then your oracle transactions, all separate and above FTP and file sharing. You can then route them out virtual circuits with specific traffic handling and shaping. The best of these devices are autonomic, discovering and linking up and creating virtual paths based on predefined policy. This is the same stuff promised in the most advanced 3GPP standards and IMS at potentially much less overhead. If service providers do not adopt and deploy this, then end enterprise customers will, rendering service providers to nothing more than being the transit pipes.
2: Single Identity: Consumers and business are sick and tired of all these separate address identifiers. Business travelers want one number that can be economically used
|
|
Perhaps the single biggest problem facing tomorrow’s service providers is controlling where all the packets go with very fine resolution. It is more than a technical problem. The technical parts are mostly solved and nearly commercial. The big problem is policy and legal boundaries |
|
anywhere in the world and the first service provider to really deliver this will win massive inflows of business. Consumers are interested in the same universal mail boxes that business service customers can have – one place for voice mails, emails, mobile messages. Combining universal addressing and integrated messaging (one number, any medium, anywhere in the world) may be the next “killer application”.
3: Viral VPN: Closed user groups were a big part of the commercial success of Advanced Intelligent Networks (AIN). MCI’s “Friends and Family” extended this concept to social networks (and billing). Today social networking is a part of the life of the youth culture and making headway in business communities via applications like LinkedIN. These social networks extend past traditional corporate and social groupings. The internet as a self-publishing medium is well established
– but with file sharing technologies proliferating into voice and video real time services, both the core provider and the web farm are being replaced by community data grids. The next, not yet here, “killer application” will be Viral VPNs, social networks, and enabling everyone as a content provider. But this product is not yet invented – or still lies in some startup’s research phase.
Looming problems: where do all the packets go?
Perhaps the single biggest problem facing tomorrow’s service providers is controlling where all the packets go with very fine resolution. It is more than a technical problem. The technical parts are mostly solved and nearly commercial. The big problem is policy and legal boundaries. Simply put, this problem meets headline today with whether access will be open, whether Skype and Vonage can be blocked by a big facilities provider. Ultimately we believe this will need to be solved by economic methods. A great thinker once said economics always wins in the long run. Diseconomies of scale cannot continue in a viable, stable marketplace. Everyone eventually must pay for transit and for content.
The best approach is to bill for traffic and to charge for traffic based on QoS priority and handling. But this is diametrically opposite of the one-price, anything, any amount pricing strategy which is current in the USA. It is also
article page | 1 | 2 | 3 | 4 | 5 | |
|
|
|
|
|
