Enter your sites, concurrent-call estimates, and WAN bandwidth. Get the codec matrix for CUCM Regions, the audio and video bandwidth to enter in each Location, and a link-utilization sanity check that tells you if your WAN will hold up during the busy hour.
CUCM Regions and Locations are the only knobs that protect your WAN from voice/video saturation. Regions tell CUCM which codec to negotiate between any two pairs of sites. Locations define how much voice and video bandwidth can be active at once over each site's WAN link — Call Admission Control (CAC). Get either one wrong and busy-hour calls start dropping or showing up at bad quality. This tool takes concurrent-call estimates, codec choices, and WAN-link sizes, then emits the Region matrix and Location bandwidth values. It also flags WAN links that can't actually carry the CAC you're asking it to.
Enter sites + codecs, click Calculate bandwidth.
Which codec CUCM negotiates between each pair of sites.
Max simultaneous audio + video bandwidth per site. Enter exactly as shown.
Tool flags if CAC + overhead exceeds WAN capacity.
Region definitions are easy: G.722 between sites with fat LAN, G.729a or Opus between sites over WAN. Most clusters get Region right. Where they go wrong is the Location audio/video bandwidth value — because nobody has the sites' concurrent-call numbers handy, engineers set it to 1,000,000 ("effectively unlimited") or to zero ("no CAC") and the site admits every call the WAN can't actually carry. At busy hour, codec retransmission and jitter drive call quality below acceptable.
The correct number is driven by concurrent calls at busy hour, not number of users. Rule of thumb for knowledge workers: 10% of phones are in a call at busy hour. For contact center agents: 80%. For call-center-heavy healthcare departments: 40%. Multiply by the WAN codec bandwidth to get Location Audio Bandwidth in kbps. Add video concurrency × video bandwidth for Location Video Bandwidth. Compare the total + 25% overhead against the WAN link capacity; if it exceeds 50% of the link, you either need a larger WAN pipe or tighter concurrency limits — or both.
One gotcha this tool protects against: Cisco's CAC is one-directional per call (a call between DAL and HOU counts against both DAL's and HOU's Location bandwidth). If you set DAL to 100 Mbps and only HOU calls are happening at 20 Mbps, you have 80 Mbps free for DAL-originated or DAL-terminated calls. The tool reports the worst-case bidirectional usage.
Opus dynamically adjusts bitrate based on available bandwidth and gives better quality at the same kbps. Cisco Collab 15 and Webex Calling default to Opus.
Set video bandwidth to 0 for audio-only sites. CUCM will deny video calls there.
SIP trunks are treated like a Location too — assign the trunk to a Location with bandwidth matching the circuit carrying it.
Enhanced Location CAC (ELCAC) is the default since CUCM 9.x — it models link bandwidth at the link level instead of per-Location. Use ELCAC for any WAN with more than 2 sites.
In a blind A/B, most users can't tell. Opus has the edge in packet-loss recovery, G.722 has better CPU/codec interop with older endpoints. Either is fine for internal.
DAL 20/4, HOU 10/2, NYC 30/6; 100/50/200 Mbps; G.729 WAN + H.264 1080p
20 users/site, 15% concurrency, video for clinical only
One 200-agent UCCX site; 80% concurrency = 160 concurrent calls; G.729
The Network & UC Health Assessment analyzes your CAC configuration, WAN usage, and call-quality metrics (jitter, loss, MOS) and gives you a bandwidth-remediation plan — which sites to increase, which to re-codec, and where to add Opus.