ProCurve Networking IP Telephony Solution technical brief

• Deploy open-standards based solutions to reduce the likelihood of massive redesign down the road.
• Provision sufficient bandwidth on WAN links.
• Prioritize voice traffic over data.
• Consider the adoption of a Service Level Agreement (SLA) with the service provider to specify network availability, throughput and average round-trip delay.

1. Perform an assessment of the current network.
   Network assessment is a valuable tool to help prepare network for IP telephony.  It’s essential to determine the ability of a company’s network to support voice traffic, so that any obstacles can be remove before deployment—for example, insufficient bandwidth or QoS settings. An assessment generates traffic and measures actual jitter and delay values across various segments of the network.  The results indicate any significant performance issues that might jeopardize voice quality on the network. A company also needs to know in advance how many voice calls its current network can safely support. In addition, a proper network baseline can assist with subsequent deployment and troubleshooting.

2. Determine the additional bandwidth and performance that will be required for new voice services.
   If the WAN will carry voice traffic, the bandwidth requirement must be assessed carefully. The link must be provisioned with sufficient bandwidth to simultaneously support the maximum number of desired voice calls and business-critical data applications. Each active telephone call typically requires 100 Kbps of bandwidth for each uncompressed (G.711) and 40 Kbps for each compressed (G.729) telephone conversation. Companies should be sure to understand the network constraints so enough bandwidth is available for worst-case scenarios.

3. Design a network topology including the IP telephony equipment
   A well-designed IP Telephony network consists of a high-bandwidth, redundant core network and an access network for edge device access. The core should have uninterruptible power supply (UPS) backup and be cross-connected in a redundant configuration, so that the failure of a single device will not result in a total network outage. The access network should provide power via IEEE 802.3af (POE) Standard and use UPS backup to provide for reliable power for E911 service.
   IP telephony deployments typically use a variety of trunks to provide PSTN connections for local calls, to ensure resiliency in the event of a WAN failure, and to provide a migration path with a legacy PBX to allow staged VoIP deployments.

4. Implement end-to-end QoS, as described in the following section.

5. Deploy telephony equipment and upgrade the network as needed.
   During deployment, you will install and configure the IP PBX(s) and additional switches as needed; configure the DHCP service for IP phones; configure the dial plan and E911 service; install IP phones; and validate phone features and dial plans.
   Consider deploying VoIP across the enterprise in phases, so you can validate your strategy before expanding the implementation.

6. Perform a post-install network assessment to validate your end-to-end solution.
   As conditions on your network evolve over time, you should continue to test the network periodically to assess the impact of new applications, increases in network usage, and any other changes that might affect voice quality.

7. Implement ongoing engineering

• Converged applications such as integrated and unified messaging, contact center applications, videoconferencing, and more

• Scalability – the ability to continue adding ports and telephones in a scalable manner

• E911 – emergency support using MAC lockdown protection to prevent unauthorized movement of phones

• A PC-based real-time communication tool that provides a single, unified interface for call handling, integrated communications and collaboration

• “Plug and work” teleworking – transparent access to the corporate voice network from any location with broadband Internet access

• ACD and automatic route selection

The reliability of communication systems is a critical requirement for enterprise business. Enterprises have come to expect 99.999 percent or “five 9s” reliability of traditional telephony systems.

To ensure that a converged network is highly reliable, these key components should be included in the network design:

• Backup and redundant power
  UPS eliminate the impact of power glitches and allows for a graceful system shutdown in the event of an extended outage. HP ProCurve PoE switches use a redundant external power supply (RPS/EPS) that minimizes the impact of a power supply failure, offering an indispensable way to provide dial tone during power outages.

• Alternate network connectivity
  Distributed resilient networks offer the ability to route around failed or inaccessible portions of an IP network. Alternate network connectivity can be achieved by using cross-connected or meshed switches in a redundant configuration, coupled with the use of PSTN trunk or analog trunk access between controllers to ensure against loss of service or call features if the WAN is unavailable.

• No single point of failure
  The ability to locate resiliency at different locations removes the concern for a total system outage due to a catastrophic event such as water damage or fire.

The efficient use of existing hardware in a distributed topology can significantly reduce hardware costs. Many redundancy solutions require dedicated expensive hardware for every controller to solve temporary and often infrequent system failures.

The inclusion of an alternate ICP in the network topology provides for call control resiliency. The alternate ICP provides services to phones in the event of a failure in the home ICP or in the event of a networking failure between the phone and the ICP. Any calls in process are maintained, and embedded voice mail continues to be provided. Once the failure is corrected, the alternate ICP directs phones back to the home ICP.

An IP telephony solution built upon the ProCurve Networking by HP Adaptive EDGE Architecture and the Mitel 3300 ICP provides for resiliency throughout the network, eliminating the risk of a single point of failure.

The ProCurve Networking by HP solution offers broad support for open standards that promote interoperability with VoIP products. Open standards allow companies to install VoIP in selected locations while maintaining their existing investment in PBX equipment. For example, the Mitel 3300 ICP integrates with any third-party PBX. It provides an incremental migration path to VoIP while preserving existing investments, with support for existing analog equipment such as fax machines, modems, conference room speakerphones and emergency 911 phones.