Industry · Enterprise communications

Voice connectivity for multi-site organisations

Larger organisations need voice that fits their network architecture, keeps local numbers at every location, survives site failures and passes procurement and security review. We plan it with your IT, network and procurement teams.

  • Central or per-site PBX
  • Local numbers per location
  • Phased ISDN/PRI migration
Glass office towers reflecting a blue sky

Multi-site architecture

Central PBX or a PBX at each site

The first architectural decision is where call control lives. Both models work with SIP trunks; the right one depends on your WAN, your sites and how IT is organised. Hybrid designs, with a central PBX and local survivability at key sites, are common.

Comparing central and per-site PBX architectures

  • Central PBX

    How it works
    One PBX or PBX cluster, in a data centre or the cloud, serves every site
    Trunks
    Central trunk capacity shared across sites
    Dial plan
    One dial plan and extension range for the organisation
    Site WAN outage
    Site loses calling unless it has local survivability
    Administration
    Changes made once, centrally
    Typical fit
    Organisations with strong WAN links and central IT
  • PBX per site

    How it works
    Each site runs its own PBX with its own trunk
    Trunks
    Trunk capacity per site
    Dial plan
    Per-site plans, linked for internal calls
    Site WAN outage
    Site keeps calling on its own trunk
    Administration
    Changes repeated per site
    Typical fit
    Sites with independent operations or weak connectivity

Local numbers

Local numbers for every location

Customers and suppliers expect to call a local number for each office, even when calls are handled centrally.

Modern glass office tower against a clear sky

With SIP, numbers are no longer tied to the building’s exchange. Each office can keep its existing local numbers, ported where possible, and the trunk delivers them to wherever the PBX runs. Outbound calls from each site can present that site’s number, so a call from the regional office shows a regional number.

Emergency calling needs particular care. In many countries the emergency services rely on the address registered against the number presented, so each site’s numbers must carry the correct address, and remote workers need a documented arrangement. Requirements are confirmed for each country before go-live.

Number availability and documentation rules differ by country. See Virtual Numbers for how new and ported numbers are handled.

Resilience

Resilient routing and disaster recovery

A disaster recovery plan for voice should state which calls go where when each component fails, and be tested.

Inbound

If the PBX serving a site cannot be reached, its numbers can be routed to an alternate destination: a second data centre, another site, or a set of mobile numbers for key staff. The order of destinations is agreed per number range.

Outbound

Outbound resilience comes from your side: diverse internet or WAN connections, a secondary PBX or SBC, and a second trunk configured on the PBX.

Testing

Test failover at go-live and at least after every significant network or PBX change. Record the date and result in your continuity documentation.

Security and governance

Security and governance controls

Voice is part of your security perimeter. These controls are agreed and documented during setup.

  • Access control

    Administrative access to PBXs and trunk settings limited to named staff, with changes logged.

  • Encryption

    TLS for SIP signalling and SRTP for media where your PBX and the route support them, agreed per deployment.

  • Calling policy

    International and premium-rate destinations enabled by policy, with channel limits per trunk or site.

  • Records

    Call detail records for cost allocation and audit, where available, and a number inventory per location.

Migration

Phased migration from ISDN and PRI

A PRI circuit carries 23 voice channels on a T1 or 30 on an E1. Replacing many circuits across many sites is best done in stages, with each stage reversible until it is proven.

  1. Step 1: Inventory

    List every ISDN, PRI and analogue circuit, the numbers on each, and what still depends on them, such as lifts, alarms and fax.

    Output: circuit inventory

  2. Step 2: Pilot site

    Move one representative site first. Confirm call quality, number presentation, emergency calling and failover.

    Output: pilot report

  3. Step 3: Parallel running

    Bring SIP trunks into service alongside existing circuits, moving outbound traffic before inbound numbers.

    Output: dual routing

  4. Step 4: Number porting by site

    Port each site’s numbers in a planned window, with the old circuits kept until porting completes.

    Output: ported numbers

  5. Step 5: Decommission

    Cease the old circuits only after live traffic has been reviewed and special lines have been re-provisioned.

    Output: closure record

Procurement and change control

Documentation for procurement and change control

Enterprise purchasing and change processes need written evidence at every stage. We provide it in a form your teams can file and review.

  • Technical description of the service, including authentication, transport and codec options
  • Configuration sheet per trunk and site
  • Itemised quote: channels, numbers, destinations and one-off work
  • Test plan and test results for each site before go-live
  • Answers to your security and supplier questionnaires
  • Data processing information, as set out in our privacy policy

Next step

Plan your multi-site voice architecture

Share your site list and current circuits. We will review your architecture, propose a migration sequence and provide documentation for procurement.

Helpful to include

  • Sites, users per site and current circuits (ISDN, PRI, analogue)
  • PBX platforms in use and where they are hosted
  • Numbers per site and countries involved
  • Continuity and failover requirements
  • Procurement and security questionnaire deadlines