Freeze one failed walk and classify its first symptom
Record one authorized call with timestamp and timezone, handset identity, SIP identity, calling direction, firmware, starting cell, movement path, intended target cell and the first visible symptom. Distinguish an established call that drops, a brief audio gap, one-way media, a handset that loses radio service, and a later inbound call that no longer reaches the handset.
Preserve B4 system status, synchronization statistics, handset state, PBX signaling and network evidence before changing placement or configuration. A later registration failure is not automatically the cause of an earlier radio interruption, and a continuous handset display is not proof that the SIP dialog or both RTP directions survived.
Measure the serving cell and radio path at the failure point
Identify the serving base before movement, the intended target base or repeater and the actual transition point. HP documents an unregistered-handset survey view and publishes reference RSSI bands ranging from excellent through disconnected. Treat those bands as vendor planning guidance rather than a universal acceptance threshold.
Repeat the route with doors, lifts, shelving, machinery, crowds and normal interference represented. The site guide recommends planned overlap and notes severe attenuation from concrete, metal and other materials. Capture the observed RSSI and cell identity along the path; do not infer handover quality from an open-space range claim or a stationary signal bar.
Prove the B4 synchronization state and chain
Record whether the system uses DECT over-the-air synchronization or the documented IEEE 1588 LAN method, the primary cell or cells, each synchronization source and the state of the affected base. Preserve the current chain, last chain change, change count and time spent in each state where the installed software exposes those statistics.
A base that is reachable by IP or serving idle handsets can still have a degraded synchronization relationship. Conversely, a reported synchronization change does not prove that the user call failed at that moment. Correlate all clocks and the exact walk before assigning cause. When multicast or peer discovery is involved, prove the required switching and routing behavior rather than treating general IP reachability as sufficient.
Separate DECT handover, SIP registration and line assignment
Mark the serving-cell transition and whether the existing DECT call continued. Then inspect the SIP layer independently: registered service, assigned handset and line, dialog identifiers, Contact, route selection and any registration change. HP documentation treats base-system registration and handset assignment as distinct from the radio path.
For a later inbound failure, follow the selected PBX Contact and INVITE to the B4 system, then the intended handset and serving cell. Do not reset every registration, delete all handsets or widen PBX permissions before proving the first missing state. Those changes can erase evidence and create a second fault.
Test codec capacity, signaling and both media directions
HP publishes codec-dependent capacity for the B4 and notes that wideband consumes more DECT channel capacity than narrowband. Record the negotiated codec, active calls on the affected cell, handset and base limits, any repeater path, PBX resources and trunk limits. A system-wide headline maximum does not establish spare capacity at the serving cell.
For the same call, correlate Call-ID, provisional and final SIP responses, RTP source and destination, sequence continuity and audio in both directions. A preserved SIP dialog, successful handover counter or received RTP stream does not by itself prove intelligible two-way audio.
Correct the first divergence and retest restoration
Correct only the first demonstrated boundary: coverage or overlap, synchronization method or source, multicast or LAN timing path, base placement, inconsistent firmware, assignment or registration, local codec capacity, SIP routing or the later RTP path. Keep rollback material and the original evidence interval.
Retest stationary calls on each cell, the same walk in both directions, one new inbound and outbound call after movement, both media directions and representative busy-hour load. Record actual results and timestamps. This workflow is documentation-derived and was not reproduced on an installed or laboratory Rove system.
Sources & applicability
Primary references for the technical details above. Operational examples and planning checklists are VoIP.info editorial guidance.
- HP Poly Rove B4 North American product and specifications ↗
- HP support: Rove B4 guides and current release documentation ↗
- HP Poly Rove DECT IP Phone Administrator Guide 8.0.11 ↗
- HP Poly Rove site planning and deployment guide ↗
- HP Poly Rove DECT IP Phones release notes ↗
Examples require adaptation to your topology. No live PBX or hardware testing is claimed.