No SIP app servers from Oracle, IBM, SAP or Sun

Bookmark Blog Post

	del.icio.us
	Digg
	DZone
	Furl
	Reddit

Ken Lee's Blog | September 19, 2005  11:30 AM | Comments (16)

Comments

• Ken, Thanks for those very interesting comments.Seems to me that Bea is miles ahead of the competition in this nascent and very importnat market. Does Microsoft has a competing product? Who do you consider as the most important competitor for WLSS? Alexander

  Posted by: agnesmuylle on September 19, 2005 at 11:45 AM

• Ken, Do you think that WLSS will get mass adoption by other non-telco verticals?

  Posted by: agnesmuylle on September 19, 2005 at 1:26 PM

• Microsoft is making tremendous investments in SIP. They have implemented the SIP stack in Windows XP, CE, MSN Messenger, and also have a server-side SIP-based instant messaging server called Live Communications Server (LCS). The SIP application programming model supported by Microsoft LCS is called Microsoft SIP Processing Language (MSPL). The IM/presence protocol supported in LCS is SIMPLE (SIP for Instant Messaging and Presence Leveraging Extensions). Currently, IBM Lotus and Microsoft are the two largest implementers of SIMPLE-based instant messaging/presence servers. I view Microsoft, along with Cisco, as the two largest distributors of SIP on the client. But Microsoft LCS is an enterprise-grade product, and likewise IBM Lotus is also targeted at the enterprise. Neither are telco-grade network platform products. WLSS is a carrier-grade SIP application server, mainly targeted towards the service provider market, and not the enterprise IT market. Technically, there is nothing to prevent the deployment of WLSS in the enterprise IT network. But because the main target market for WLSS is the telecom service delivery network, the main competitor to BEA in this target market is a 150-person company called Ubiquity Software. Ubiquity isn't a big threat to BEA WLSS just by themselves. It is through their strategic alliances with large IT/SI vendors like IBM Global Services, HP, and with large NEPs such as Lucent, Nortel, Siemens and Nokia that Ubiquity becomes a global competitor to BEA WLSS. In terms of whether or not BEA WLSS can gain mass adoption in non-telecom industries, and within enterprises, I believe the answer lies in the understanding of potential demand for real-time, interactive, multiparty, multimedia communication and collaboration services in industries outside the telecom industry. Regardless of whether the end-user is consumers, or business users, the concrete benefits provided in terms of productivity gains, enhanced customer satisfaction, or more efficient and effective communications, the many as-yet-uncommercialized IP-based services that a technology such as Session Initiation Protocol (SIP) can enable will revolutionize the way in which all of us use, access, and interact with the Internet. So in a nutshell, I would say the market for WLSS, in all industries, is massive in the next 2-3 years.

  Posted by: jusmagk on September 19, 2005 at 2:30 PM

• Thanks for your insights. One has to wonder why Bea doesn't have those strategic alliances with NEPs....but it seems Bea is actively pursuing partnerships with those NEPs. In my opinion reseller agreements with those NEPs are absolutely key for mass adoption of WLSS in telco and non-telco markets.

  Posted by: agnesmuylle on September 20, 2005 at 1:05 AM

• There is a lot of work needed to do interoperability testing (IOT) with large NEPs, such as Ericsson, and others. Also, the telco market is very regional, and the NEPs who dominate those regions are also different. For instance, in Japan, NEC dominates, whereas in Korea, it's Samsung. NEPs are very far ahead in their product lines around SIP and IMS, as well as standards activities. As such, as a new entrant, BEA has a lot of ground to make up with the NEPs. Each NEP has a SIP application server strategy, and for some, reselling is preferred, whereas others it's OEM, or reference selling. Yes, NEP partnerships and IOT will be one of the strategic levers in selling a SIP-IMS application platform into the telecom service provider markets. In terms of non-telecoms, it's more of an enterprise IT play, and I don't see telco-centric NEPs having a lot of influence there. Instead, enterprise-centric NEPs, like Avaya and Cisco, will have a lot of success and influence in the non-telco accounts. - Ken

  Posted by: jusmagk on September 20, 2005 at 9:26 AM

• Ken, correct me if i'm wrong but basically the primary competitor for WLSS is the proprietary SIP solutions of the NEPs, both for telco and non-telco markets?

  Posted by: agnesmuylle on September 20, 2005 at 9:37 AM

• Ken, can you elaborate those IOT with NEPs? thanks!

  Posted by: agnesmuylle on September 20, 2005 at 9:55 AM

• I want to make sure we differentiate between the SIP products NEPs provide, versus what an IT vendor like BEA provides.

  NEPs provide SIP network servers, which are logical SIP network elements specified as part of the IETF SIP standard. These are SIP Proxy Server, SIP Registrar Server, SIP Redirect Server, and SIP User Agent (client and server). Often these SIP network servers are co-located in a single piece of hardware, and productized as a SIP Softswitch.

  What BEA and it's competitors provide are SIP application servers, which connect to a SIP network via the SIP Proxy Server. There are several SIP application programming models, but the Java-based model has been standardized as SIP Servlets (JSR 116). Other models include Call Processing Language (CPL), SIP-CGI, and Microsoft SIP Processing Language (MSPL). But in general, SIP application logic can be developed using any programming language.

  Many NEPs have made the strategic product decision to either include a SIP application server, along with their SIP network server product lines, or have decided to partner with one of BEA's SIP application server competitors, such as Ubiquity. But there are a few global SIs who have also partnered with BEA's SIP application server competitors, and are actively reselling or reference selling a SIP application server. In that regard, BEA comes into competition with an NEP or SI.

  But in general, BEA WLSS partners and integrates with NEP's SIP network server product lines, and does not compete with NEPs or SIs. As I've mentioned, NEPs main strengths are in the telco's, and not in the enterprise IT market, which is where non-telco usage of SIP network and application servers reside.

  - Ken

  Posted by: jusmagk on September 20, 2005 at 10:02 AM

• Thanks. If I inderstood you correctly, some NEPs have built their own SIP application server?

  Posted by: agnesmuylle on September 20, 2005 at 10:24 AM

• Not anymore. Some NEPs did start down that path, but most have discontinued their efforts, and have decided to partner in the SIP application server product segment. Instead, they've decided to focus on the core SIP and IMS network servers.

  This provides a golden opportunity for SIP application servers, such as BEA WLSS, to become the NEP's SIP application server solution. Depending on the regional market, there are many small ISVs who supply SIP application servers, whether it's based on CPL or SIP Servlets, or both.

  - Ken

  Posted by: jusmagk on September 20, 2005 at 1:57 PM

• Yet, I might be wrong but most of the WLLS' competing ISVs can't provide the scalability, relibaility and platform approach that WLLS provides...

  Posted by: agnesmuylle on September 20, 2005 at 3:08 PM

• Yes, that's correct. And WLSS is the only SIP application server on the market that's integrated with a J2EE container, providing the inherent benefits of J2EE services to SIP applications.

  Posted by: jusmagk on September 20, 2005 at 4:00 PM

• Ken, I read a lot about SIP and IMS enabling new services as well as the blending of services such audio, video and text but I have not found anything yet on enterprise services. Enterprise application vendors such as SAP, Oracle, MS and RIM offer proprietary solutions to enable mobile access to their enterprise applications. Can IMS help to provide standard solutions? Do I understand it right that, at least in theory, if a vendor would IMS-enable its enterprise service and I had the appropriate IMS enabled client, I would not need a proprietary solution anymore? An example would be MS Exchange IMS-enabled and an Exchange client on a IMS-capable handset. Could the IMS solution provide a standard-based equivalent to a Blackberry server and handset? SAP and Oracle are currently not providing a SIP Server. Could I use the BEA SIP server to connect to SAP, Oracle and other enterprise apps? I assume that I have to have a IMS-enabled mobile client for those and a counterpart on the vendor's app server side. Could a third party, such as Bea provide a software layer in between for the client and the server to enable this type of communication using IMS and is that likely to happen? And finally, what are the handset requirements for IMS? Do I just need a 2.5G or better handset with a runtime/OS (Java, Brew, Symbian) on it that allows me to run the appropriate client software? Would the SIP part be handled by the client app or would I need an extra software stack on the device? I hope the questions make sense and I'm sorry for the long post - thank you.

  Posted by: renatos on October 17, 2005 at 5:47 PM

• IMS is a telco standard for deploying an IP-based next-gen network architecture. It is based on existing Internet standards, such as SIP, RTP, and Diameter, among others, but I don't see how an IMS-based telco network could get deployed within an enterprise network. And I'm not sure there can technically be an "IMS client", since an IMS network is access agnostic. Different telecom network can access IMS networks via media gateways, or GPRS/UMTS gateways. SIP devices are only one type of device that can access IMS services.

  I think the question you might be asking is one about SIP in the enterprise, since deploying SIP into an enterprise network, or call center, or portal are all very possible scenarios. In this case, the client, or the SIP User Agent Client (whatever device that may be), will need to have a SIP protocol stack on it, or otherwise, it will not be able to participate in SIP communications. Of course, non-SIP devices can communicate with SIP clients through media gateways.

  - Ken

  Posted by: jusmagk on October 17, 2005 at 10:13 PM

• Ken, thank you for your clarification. The benefits of IMS seem to center around fix-mobile convergence and the blending of multimedia services. In the wireless sector most of the focus currently is on the provisioning and blending of services such as SMS, MMS, Video Messaging and PoC (Push-to-talk over cellular). Carriers seem to also benefit from the possibility of a single billing system. Wireline carriers could benefit from a smooth transition from traditional phone service to IP telephony by providing a network that can cope with both, depending on the end user device and service access. If the user picks up a landline, he would get connected via gateways to the IMS network that would transform the analog signal to IP for transport to the other endpoint, which could be a user using an IP telephone or correspondent software on another device (even mobile phone). Wireless carrier would have the same benefits but it's likely to take longer for them to transition to an all IP network and it would canibalize their current revenue stream. Do I understand these benefits correctly? To my understanding, it is a framework that if adopted, it would allow a best of breed selection for application vendors for any given service. Could you give me some concrete application examples that would benefit from an IMS framework? Do you know if Oracle, IBM, SAP and Sun have plans to support IMS and either release SIP clients and/or a SIP app server for the IMS framework? When you say that SIP devices are only one type of client to access IMS services, which other types can you list that can connect directly w/o having to go through a media gateway? Thank you again for helping me understand the subject better.

  Posted by: renatos on October 20, 2005 at 12:59 PM

• What's the need to have an IMS server if you don't have proper clients? I just bought an EV-DO enabled handset from Ubiquam. (As far as I got it's based on Qualcomm's chipset). They use a perversed modification of java environment- the so called BREW. Now it's the mid of 2007 and they still seem to have no SIP support on this platform. Just imagine this! The only commercial mobile network with datarates capable to support live SIP calls (not just text messaging) - and there are no handsets or applications to do so! incredible! now they are trying to flog their customers this iptv bullshit. So my question is as following: are there any plans on developing the SIP platform on BREW? Whom to ask? I do not want to build my own business (although, why not?)) - I'd like to use this sorta client just for myself.

  Posted by: Red_Is on July 20, 2007 at 7:38 AM