Spring / Summer 1984

by Carlyle Smith and Allan Hogg LES5.4. 1

System X exchanges will provide a range of additional facilities for customers connected to

them. Despite British Telecom's plans for the rapid introduction of System X (digital)

exchanges into the network, there will still be a significant number of lines connected to non

digital (analogue) exchanges well into the 1990s.

For this reason British Telecom embarked upon a programme for the enhancement of

analogue exchanges to provide the top eight facilities- known as Star Services. A pilot

scheme for analogue enhancement was launched in 1982 when a number of firms were

invited to tender for the supply of suitable equipment. Two manufacturers, TMC and IBM,

were selected, and each contracted to supply equipment for enhancing 50,000 lines in some

200 analogue exchanges.

The replies to the tender suggested a number of approaches to analogue enhancement.

However, both the TMC and IBM equipment's can be described loosely as 'front-end black

boxes'. These 'black boxes' are processor controlled switches which are connected

between the customers' lines and the host exchange line circuits via the MDF and IDF.

Figure 1 shows the arrangement

The eight Star Services being provided were chosen after consideration of the marketing potential and ease of implementation the services are :

Code Calling

Repeat Last Call

Charge Advice

Reminder Call

Call Diversion

Three way Calling

Call waiting

Call barring

The IBM SSU consists of a service access unit (SAU) and an administrative control unit (ACUL as shown in Figure 2. The SAU has two identical and independent switching sub-units, each having a 480-port microprocessor-controlled non-blocking switch. One hundred and eighty ports are used for terminating lines from customers, and a similar number are used for connection to customers' calling equipment on the host exchange. All outgoing calls are made through a pool of 35 lines on the host exchange. Incoming calls are completed by connecting the customer's exchange appearance to his line appearance. The two sub-units share a recorded announcement system (RASL a common - 50V DC supply for feeding line current, a ringing voltage supply and the interface to the ACU. 

The ACU is an IBM Series 1 minicomputer, and provides backup storage, maintenance facilities, and data input and retrieval facilities. The backup storage medium is a 1 megabyte diskette, and holds- in addition to certain data and programming information relating to the Series 1 -call records, exchange data, and customers' facilities data. Communications with the ACU is through a VDU and keyboard, and hard copy can be obtained from a printer. 

Charging data from IBM SSUs is collected by a central data processing system, sometimes called a billing data collection unit (BDCU). The BDCU will poll each of the 139 IBM SSUs once every hour using the pstn. Output from the BDCU is on 112 inch magnetic tape using a cartridge format; it is therefore suitable for direct input into CRACS. 

The BDCU also provides a further level of backup for Star Services data, such as customer activation status, time reminder service calls 

and so on. In the event of a diskette failure the SSU will call on the BDCU to 'down-load' the data. 

The TMC SSU, like its IBM counterpart consists of a switch- the Star Services Exchange (SSX), and a control computer- the peripheral control system (PCS). The arrangement is shown in Figure 3. The SSX has 12 identical and independent switching sub-units, each capable of handling 48 Star service lines. Each 48-line module has its own microprocessor, announcement machine, outgoing pool lines and so on; but they all share the alarm card, and, of course, the PCS. While all outgoing calls from the IBM SSU use pool lines, the TMC SSU steers its outgoing calls either over pool lines, or the customer's exchange appearance, depending on the Star Services which are activated when the call originates.

The PCS is a Data General Eclipse minicomputer, with Winchester disk as well as diskette for backup storage. It performs the same main functions as the IBM equivalent; however, call charge data is stored on magnetic cartridges, which are sent to CRACS for processing.

SSUs can generate call charging data for those customers connected to them. This data can be used for the production of itemised statements 

as well as for obtaining the bulked value of call units used. Itemisation of calls is not being offered as part of Star Services; but the itemisation data will be used to provide customers with a summary of Star Service call charges. The mass storage medium for call charging data is magnetic cartridge, or magnetic tape using a cartridge format The tapes or cartridges produced by SSUs are processed at the Cartridge Reading and Control Centre (CRACS), which produces an input suitable for one of the four billing computers which produce telephone bills. CRACS also processes System X call charging data, and by using the System X formats for SSU data, a common software package can be used at CRACS. Although the IBM and TMC Star Service Units (SSUs) both use the 'front-end black box' philosophy, there are important differences between them. The following descriptions h1ghlight these differences.

The basic approach adopted is to diagnose faults down to PCB level on-site, change the faulty parts, and repair off-site in Area Repair Centres (ARC) where possible. There will inevitably be some cases where the SSU faults are beyond the ability of the exchange maintenance staff to cure. To cover this aspect regional back-up groups are being trained, and a fault escalation procedure will operate in stages from exchange to Region, Region to THO and THO to prime contractor. To implement th1s philosophy British Telecom HO has arranged: 

• For the training of staff to an appropriate level for their particular function, that IS, Tier 1: for on-site maintenance staff so that the equipment can be operated efficiently and faults located and cleared speedily Tier 2: for Regional back-up staff who will have a second training course designed to provide them with a greater in-depth knowledge of the system. They may also have specialist test gear for locating obscure faults. 

• For spares to be held On-site: generally speaking those cards which, if faulty 1n service, could cause loss of service to some or all of the customers connected. Consumable Items are also held. Off-site: generally those cards which, if faulty in service, would cause a degradation of customer service. 

• A 12-month warranty for the equ1pment. Dur1ng this period, faulty boards will be returned to the contractor by Regions, or ARCs, for a free exchange. Both British Telecom and the prime contractors will want to Inspect any faulty equipment to look closely for any des1gn or component problem that may appear, and to check the system's reliability 

Arrangements for the repair of any faulty equipment after the expiry of the warranty period differ for each contractor. For faulty IBM boards, the present policy is not to repair. but to replace with new ones. However we are still negotiating with IBM for a change to the present arrangements, thus enabling British Telecom to repair faulty cards. In the TMC case, the Company are providing British Telecom with all the necessary information to enable the ARCs to undertake the normal repair functions, and the new Computer Aided Record System (CARS) will be used to check failures. The performance of the enhancement systems will be closely monitored in the early stages of customer service. Statistics generated by the SSUs, the SSU exchange fault log and customers complaints will be used to:

• Monitor traffic patterns 

• Monitor the use of the individual services for marketing 

• Check the equipment provisioning standards 

  Monitor the reliability of the equipment. and 

• Highlight any design problems and component failures that occur. 

Initially, these statistics will be collected monthly, but will be reviewed in the light of experience gained and extended when the time is Judged to be right. 

The allocation of sites between the two contractors was made on a regional basis. London is shared between both companies; TMC has for the four southern most English provincial regions, and IBM the remainder of the UK. Installation of SSUs started in June 1983 and was due for completion early this year. At the time of writing 5 units have been handed over by the contractors in Cheltenham, Stroud, Chippenham, Clerkenwell and Belgravia. Thus we are already providing a significant number of our customers with a range of services, that would normally be available only on processor controlled exchanges, like System X. (01-432 2456)