Convergence faces

Greenfield convergence of voice/data/video to an all-IP infrastructure in local access loops and campuses will begin in 2001 and continue through 2004, but the carrier core will gradually transform to IP through 2006. Users will continue enterprise pockets of convergence (toll bypass, teleworker, PBX replacement/augmentation, etc.) during 2001-03, but security concerns, poor quality, unrealized savings, etc. will prevent widespread enterprise adoption before 2005.

Through 2006, campus migration toward a converged IP network (for voice, video, and data) will remain limited due to a lack of demonstrable ROI and elegant management tools, multiple internal enterprise directory instances, and limited user pull. Greenfield and branch-office locations will be initial adopters (2002-04), but convergence will suffer some market stagnation in 2002/03 (with the exception of carrier cores and international toll bypass). Voice and data over wireless will remain a niche, with multi-band devices, handsets, and adapters emerging in 2002, and voice over wireless LANs appearing in small or temporary offices by 2004.

Enterprises will increase videoconferencing use, with some IP acceleration for cost savings. In 2002/03, desktop videoconferencing adoption will remain limited as the endpoints commoditize and intelligence moves closer to the OS (e.g., Windows XP).

Visions of an all-IP infrastructure will hit the spotlight in 2002, accelerated by success stories of IP telephony (campus deployments of fully featured IP telephones) and voice over IP (VoIP; OSI layers 1-4, packetized voice in transport-only form) implementations in high-profile enterprises. Initial deployments will be fueled by the press and, by extension, the enterprise, with anticipation. Tough economic times accelerated investigation and implementation of IP wide-area network solutions to include voice and data over point-to-point private circuits (i.e., frame relay) as well as IP virtual private networks (VPNs) as a cost-saving tactic, with measurable success. In fact, through 2002, approximately 50 per cent of Global 2000 organizations have implemented some type of VoIP or IP telephony. These implementations have illustrated many of the complexities that occur when introducing additional (and somewhat unfamiliar) applications to the enterprise infrastructure; quality-of-service (QoS) requirements, the necessity of new skill sets, and an increased focus on infrastructure management have followed.

Within the next 18-24 months, IP telephony will be at par with traditional voice infrastructure counterparts (e.g., PBX, key systems, centrex) from a feature and functionality standpoint (e.g., attendant consoles, call trace), and reliability can be engineered into the hardware and applications through redundancy and backup resources. However, as enterprises continue proof of concept and subsequent rollouts of this technology, the expected ROI is currently making a poor showing, making replacement of currently functioning systems impractical.

Upgrading or tweaking the existing local-area network (LAN) will lead to the following expenses, which are required in almost every implementation but often not considered:

    Building complete redundancy in the core, which is necessary in all cases due to the increased criticality of the LAN infrastructureIntensified monitoring and managementBackup power in wiring closets Additional QoS requirements

QoS processes and products remain specific to pure-play, single vendor implementations, limiting the effectiveness of heterogeneous environments, which is the ultimate goal of standards-based single-protocol applications.

Many enterprises, recognizing the associated cost savings with downsizing office space, are encouraging telecommuting as a choice to employees who are seeking quality-of-life influences. Concurrent broadband deployment options within the local exchange carriers have increased the viability of “extending the enterprise” to teleworkers (the primary convergence application identified by a recent META Group multi-client study, Campus Convergence 2001: Trends and Issues). This will continue throughout 2004 with improved delivery of enterprise applications (e.g., telephony, distance learning, enhanced collaboration) to telecommmuters. Wireless LAN options (currently employing 802.11b) have also commanded attention as an augmentation to wired LANs. Two separate META Group multi-client studies indicated that more than 70 per cent of the respondents have or plan to deploy wireless LANs in 2002.

Historically, these have been targeted in such verticals as retail and distribution, but increasingly are being considered to accommodate temporary offices (less than two years) as well as hoteling solutions (providing office space for mobile or teleworkers in corporate facilities). Wireless LANs will continue to support niche applications through 2004.

Corporate directories, not significantly impacted by voice applications previously, become an even more integral pivot point for the enterprise. Use of a common directory with IT systems to administer and assign users to resources (e.g., telephones, office space, MAC addresses) – coupled with a common set of permissions (referred to in the voice world as dialing restrictions and class of service [features], and most often found in databases) – will widen scope of the corporate directory, making it necessary to expand directory integration to include the user self-service functionality promised by converged platforms. The various vendor types for both directory and converged systems will complicate this.

Carriers continue to upgrade switching platforms to IP for better efficiency, though the upgrade cycle has been hampered by the general downturn in the telecommunications sector. Encryption requirements and poor network delivery characteristics of Internet-based IP VPNs will remain a barrier to voice services in the public arena. Instead, network-based VPNs (based on MPLS technology) will provide the required network services (e.g., mesh topologies, differentiated services with low latency/jitter, low packet loss), maturing during 2002/03.

Business Impact: Providing additional value via cost reduction is imperative for any technology transition considered during 2002, and likely 1H03, unless there is a compelling, short ROI, or business case (e.g., competitive advantage).

Bottom Line: Transition to a converged infrastructure is far reaching within the enterprise technology portfolio. Enterprises that have moved past the initial phase of lab and performance testing must evaluate the return on investment and management expectations to establish success criteria.