506 lines
27 KiB
Plaintext
506 lines
27 KiB
Plaintext
Signalling System 7 (SS7)
|
||
|
||
Whether a call is made to the phone in the house next door or on another
|
||
continent, it becomes part of traffic on a network called Signalling System 7,
|
||
or SS7. Over the past five to ten years, telephone operating companies have
|
||
been upgrading their networks to use this standard communications protocol,
|
||
providing them with faster call setup times and the ability to expand their
|
||
service offerings. There is a proliferation of communications services ranging
|
||
from Caller ID to cellular service to ISDN and the forthcoming AIN or Advanced
|
||
Intelligent Network. SS7 plays a major part in many of these services
|
||
providing the means for transporting information between locations.
|
||
|
||
In-band vs. out-of-band
|
||
|
||
When a phone call is made, call-control information is sent to the locaal
|
||
telephone office. The digits dialed are the main routing components that
|
||
determine a call's destination. If the dialing is for a local call, the call
|
||
may be connected from the same office from which the originating line
|
||
terminates. The telephone switch at the local office that services a phone
|
||
line may have to route a call to another office connected by a trunk. Call-
|
||
control signals such as the number dialed and the answer indication from the
|
||
other end are information used for managing a call connection.
|
||
|
||
Using traditional signalling methods, the trunk between the two offices
|
||
carries information down the same set of wires that the voice signal travels.
|
||
This is called in-band signalling because the call-control signal is sent down
|
||
the same path as the voice signal. SS7 handles all these tasks on a separate
|
||
facility know as a signalling link. The signalling link can handle the call-
|
||
control information for many calls going on simultaneously. The actual voice
|
||
path between the two offices is still over the trunks, while the call-control
|
||
signalling is traveling on a separate communications channel. This is called
|
||
out-of-band signalling.
|
||
|
||
SS7 is essentially a packet switching network. Signalling information is
|
||
carried in data packets between the telephone offices in much the same manner
|
||
as X.25 or other packet switching protocols previously installed. This packet-
|
||
switching network is overlaid on top of the existing telephone network, adding
|
||
an entirely new diminsion. This gives the telephone network a number of
|
||
advantages over the traditional signalling system. The primary benefit is
|
||
increased bandwidth for call signalling. The voice trunk is limited since its
|
||
primary responsibility is to carry voice or data. SS7 provides additional
|
||
bandwidth, a standardized protocol for sending information between different
|
||
vendor equipment and increased data transmission speed.
|
||
|
||
Demand for network services
|
||
|
||
The greatest benefit for both the telephone operating companies and their
|
||
subscribers is the increased capability to provide network services. Prior to
|
||
using SS7, many telecommunications equipment vendors had proprietary means for
|
||
sending feature-related signalling between offices. This prevented true
|
||
networking of services. When Integrated Services Digital Network (ISDN) was
|
||
introduced into the marketplace a few years ago, one complaint was its limited
|
||
service across geographical locations. This created situations which came to
|
||
be known as "ISDN islands." SS7 eliminates this problem by encapsulating the
|
||
ISDN call information in packets and transporting them across the network,
|
||
bridging the islands.
|
||
|
||
SS7 enables or enhances a number or services including:
|
||
|
||
o Enhanced 800 service
|
||
o Custom Local Area Signalling Services (CLASS)
|
||
o Advanced Intelligent Network Services (AIN)
|
||
o ISDN Connectivity
|
||
o Cellular Service
|
||
|
||
Until recently, when you purchased 800 service, the number you were given
|
||
actually belonged to the local company that was the service provider. FCC
|
||
rulings in recent years haved changed this scheme. Now with an enhanced form
|
||
of 800 service, the 800 number can be retained by the subscriber even when he
|
||
switches service providers. However, this means that the telephone company can
|
||
no longer determine which service provider to route the call to just by the
|
||
800 number that's dialed. An SSP uses TCAP (Transaction Capabilities Part) to
|
||
query a database at an SCP to determine the service provider for routing the
|
||
call to as well as other information associated with the call. An SSP running
|
||
CLASS uses TCAP to exchage information on the availability of a called number
|
||
with another SSP.
|
||
|
||
Custom Local Area Signalling Services (CLASS) use SS7 capabilities to deliver
|
||
services such as caller ID, automatic redial, and call screening. Call
|
||
screening allows the consumer to selectively accept or reject calls from
|
||
selected numbers. The information for these services is transported between
|
||
offices via SS7 packets.
|
||
|
||
The standards
|
||
|
||
The ability to provide information between phone offices without regard for
|
||
which vendor's equipment is used requires global standards. Standards are
|
||
developed at different levels by different organizations. Global SS7 standards
|
||
are developed by the International Telecommunications Union Telecommunications
|
||
Standardization Sector (ITU-TS), formerly known as CCITT. Different countries
|
||
make their own refinements of the ITU standards as necessary. The discussion
|
||
here is limited primarily to North American networks. The American National
|
||
Standards Institute (ANSI) and Bellcore further refine the ITU standards for
|
||
North American and Regional Bell Operating Companies (RBOCS) respectively.
|
||
|
||
Virtually anyone in the communications field today will recognize the Open
|
||
Systems Interconnection (OSI) model. The OSI stack was developed by the
|
||
International Standards Organization (ISO) and contains seven layers
|
||
identifying communications functions between two nodes such as the physical
|
||
medium used for the connection, the error correction method, addressing
|
||
scheme and so on. SS7 is also a protocol and is based on the OSI protocol
|
||
stack.
|
||
|
||
The Protocol
|
||
|
||
The SS7 protocol (refer to Fig.3) is composed of:
|
||
|
||
o Message-transfer part
|
||
o Signalling connection control part
|
||
o ISDN user part
|
||
o Transaction capability application
|
||
|
||
The MTP (message-transfer part) provides the basic transport system for all
|
||
SS7 messages. It is responsible for getting information from one network node
|
||
to another in a reliable fashion. It makes up the first three levels of the
|
||
protocol stack: the physical, link and network layers.
|
||
|
||
Layer 1, the physical level, specifies the actual medium used for
|
||
transmission. It uses a four-wire connection and typically a bit-rate transfer
|
||
of 64 kilobits per second (kb/s) or 56 kb/s. V.35 connections may also be used
|
||
with incremental transmission rates up to 64 kb/s.
|
||
|
||
Layer 2, the link level, provides a number of functions to ensure that there
|
||
is a good connection between nodes for communicating. Error detecting, error
|
||
correction, signalling unit alignment and signalling link alignment are all
|
||
part of the link layer's responsibility. It is at this layer that the actual
|
||
signalling unit is formed. Signalling units are simply SS7's version of
|
||
packets. Signalling units are transmitted across the signalling link
|
||
continuously whether there is any information to transmit or not. When there
|
||
is actually a message to be sent, it is sent an MSU (message signalling unit).
|
||
During periods when there is no inoformation to send, FISU (fill-in signalling
|
||
units) are sent. This continuous stream of packets ensures that link problems
|
||
are detected immediately. There is a third type of signalling unit, an LSSU
|
||
(link-status signalling unit) which is used to convey changes in the status of
|
||
the link between the two ends.
|
||
|
||
The SCCP (signalling-connection control part), which is part of Layer 4,
|
||
provides additional routing and network management functions to the MTP. It
|
||
allows applications to talk to each other at different nodes and it provides
|
||
network management capabilities at the application level. For example, an
|
||
application may want to re-route a message in the event of an application
|
||
failure. You'll note from the SS7 protocol model (Fig.3) that there is a
|
||
connection between the ISUP and SCCP layers. SCCP contains connection-oriented
|
||
procedures that may be used by ISUP; however, ISUP doesn't use them today. It
|
||
can communicate directly with MTP which suffices for current ISUP needs. New
|
||
services may however make use of the SCCP connection-oriented capabilities.
|
||
|
||
The ISUP (ISDN user part) of Layer 4 provides connection-oriented signalling
|
||
between nodes. This type of signalling relates to setting up, taking down and
|
||
monitoring the connection of the actual voice path between offices. ISUP is
|
||
what provides the capability for phone calls to be completed. It also provides
|
||
services such as Caller ID. The name ISDN User Part can be a bit misleading,
|
||
however, because you don't need to have ISDN to use this capability. It was
|
||
however designed with ISDN capability in mind.
|
||
|
||
TCAP (transaction capabilities part), also part of Layer 4, allows
|
||
connectionless communications between two applications using a generic
|
||
language. It provides query and response capabilities allowing nodes to
|
||
request and respond to network and service information regardless of whether
|
||
there is an actual call established between offices. This opens up an entire
|
||
world of database interaction allowing centralized network intelligence in
|
||
handling calls.
|
||
|
||
As mentioned earlier, SS7 is essentially overlaid on the existing telephone
|
||
network. This introduces some new network elements as well as giving additonal
|
||
capabilities to previously exiting ones. The network is made up of a number
|
||
nodes called signalling points. Figure 4 shows a network example consisting of
|
||
connected nodes.
|
||
|
||
The SSP (service-switching point) is the telephone office with SS7
|
||
capabilities. It can originate and terminate messages but cannot transfer
|
||
them. The STP (signalling-transfer point) takes care of the transfer part. It
|
||
is the message-switching hub of the network, essentially a big packet switch.
|
||
Many of the routing decisions are made at the STP. Without this node, every
|
||
SSP would need to have a connection to every other SSP it was required to send
|
||
messages to. This would quickly grow into a complicated scene. STP's are
|
||
usually deployed in mated pairs to provide redundancy.
|
||
|
||
The SCP (service-control point) provides database services. Telephone offices
|
||
can send queries to the database requesting information regarding 800 numbers,
|
||
Private Virtual Network numbers and calling-card numbers, to name a few.
|
||
|
||
Network routing
|
||
|
||
We have seen that the physical connection between offices that provides the
|
||
signalling communications is called a signalling link. This link is actually a
|
||
part of a linkset. A linkset is simply a set of signalling links connecting
|
||
two offices. ANSI specifies that a linkset may contain up to 16 links. Many
|
||
offices may be able to handle all of their traffic on a single link per
|
||
linkset. However, the desire for additonal traffic capacity or just alternate
|
||
facilities in the case of a facility failure often merits additional links.
|
||
There can only be one linkset defined between two offices. While links define
|
||
physical connections between offices, a route describes the path between a
|
||
node and a destination.
|
||
|
||
A route may consist of multiple linksets. There may be several routes from one
|
||
node to another. Each route follows one or more linksets to its destination.
|
||
Just as there may be several links in a linkset, a routeset is a set of routes
|
||
which describes alternate paths from a node to a destination. When a node
|
||
needs to send a message to another node, it chooses a routeset which is
|
||
associated with a destination, then chooses a route within the routeset
|
||
(remember that a route really just describes a linkset), then chooses a link
|
||
within the linkset.
|
||
|
||
That brings us to the next topic of routing: how to determine which node to
|
||
send a message to. Every office is assigned a point code. This is the address
|
||
of the office, simply a number to uniquely identify it. Point codes vary in
|
||
format depending on the country and the standards they use. The ANSI standards
|
||
used by North America designate a 9-digit point code to identify each node in
|
||
the network. Each message contains both a destination and point code to
|
||
identify the office to send the message to and an origination point code to
|
||
identify the office sending the message. Within each office, translations are
|
||
done to map this address to a routeset for which outgoing messages are to be
|
||
sent. This means that each node must designate routesets for each pointcode it
|
||
wishes to directly send messages to.
|
||
|
||
The decisions about how routing will be done can vary from company to company
|
||
and are made by administrators of the network. This type of routing based on
|
||
the point codes is done at the network level of the MTP and its primary
|
||
responsibility is getting messages from one node to another.
|
||
|
||
The next level of routing to consider is routing to an application, or in SS7
|
||
terms, a subsystem. Subsystem routing is also based on a number designated for
|
||
a specific application. This number must be agreed on by different companies
|
||
so that a subsystem number identifying a particular subsystem can be
|
||
interpreted correctly. These are usually not defined in the more general
|
||
standards, but are usually defined by those involved in network
|
||
administration. For instance, Bellcore, the research and developement
|
||
organization for the regional Bell operating companies has defined a number of
|
||
subsystems for their clients in the US. One example is Custom Local Area
|
||
Signalling Service (CLASS), which has been defined as subsystem 251.
|
||
Therefore, two offices sending CLASS related messages would designate a
|
||
subsystem of 251 in the message. Subsystem routing is the responsibility of
|
||
the SCCP level of the protocol. At the beginning, we determined that the
|
||
digits of the telephone number played a major part in determining how your
|
||
call is routed through the network.
|
||
|
||
One of the popular buzzwords in SS7 terminology is something called global
|
||
title translations. A global title is simply a set of digits. These may be
|
||
digits dialed by a subscriber or provided by an application by some means.
|
||
Global title translations is the process of mapping those digits to an SS7
|
||
address, namely a point code and a subsystem. We've determined that a point
|
||
code can route a message to an office and a subsystem number can route to an
|
||
application. Once these two pieces of information are determined, we have the
|
||
means to get a message from our application to an application somewhere else
|
||
in the network. Traditional routing in the telephone network is based on
|
||
digits. You realize that fact every time you pick up the phone. However, the
|
||
SS7 network routes its messages based on the point code and subsystem.
|
||
Therefore global title translations are needed, which is also a function of
|
||
the SCCP layer of the protocol.
|
||
|
||
Let's summarize how messages are routed across the SS7 network. When a call
|
||
begins its routing process, the dialed digits are examined. For connection-
|
||
oriented calls using the ISUP layer of the protocol, the digits are mapped
|
||
internally to the appropriate point code by the sending the message to the
|
||
next node. The ISUP message also contains a circuit-identification code to
|
||
identify which trunk the message relates to. This is necessary because it will
|
||
be traveling on a different facility from the actual voice or data call. If
|
||
level 2 has determined that both ends of the signalling link are at a suitable
|
||
level of service, level 3, the network level, routes the message to the next
|
||
office based on the point code.
|
||
|
||
Now, assume that you're sending a TCAP message to a database to determine
|
||
information related to an 800 number. (Refer to Fig.5) The point code to send
|
||
the message to would still have to be determined, but a subsystem number would
|
||
be needed also. The protocol model shows that a TCAP message must ride on top
|
||
of the services of SCCP. Since TCAP is a connectionless message that's
|
||
normally related to an application, the subsystem routing service of SCCP are
|
||
needed. This is where global title translations comes into play. From the SSP,
|
||
the message might be sent to the STP to let it perform translation on the 800
|
||
number and determine how to route it to the database. In fact, this is what's
|
||
normally done.
|
||
|
||
It is not necessary for all of the offices to have knowledge of the database
|
||
locations. This can be taken care of at a centralized point, the STP. Routing
|
||
might occur through multiple STPs before reaching the SCP, but by the time it
|
||
arrives, the final point code and subsystem have been determined so that the
|
||
800 application software at the appropriate database can handle the message.
|
||
|
||
The self-healing network
|
||
|
||
The headlines citing major SS7 outages give insight into the importance of the
|
||
signalling network. If an office uses SS7 signalling, its loss means that the
|
||
office can't communicate with the rest of the world. It becomes isolated. The
|
||
network and protocol design take this into account, providing alternate
|
||
routing, compulsive restoration where possible, and internodal communications
|
||
to coordinate activities concerning degradation or loss of service. The
|
||
network management implemented by the MTP can be divided into three
|
||
categories: signalling-link management, signalling-route, and signalling-
|
||
traffic management.
|
||
|
||
Together, these management procedures attempt to maintain service by re-
|
||
routing or controlling traffic when there is congestion or a failure in the
|
||
network. Built-in recovery procedures attempt to restore network components to
|
||
service if possible.
|
||
|
||
Signalling-link management is responsible for maintaining the path between
|
||
nodes. If excessive errors are detected by the link layer, the link may be
|
||
deactivated. Siganlling link management will attempt to restore the link
|
||
through a process known as signalling link alignment. This involves an
|
||
exchange of signalling units (LSSUs) to bring the link back to the proper
|
||
state. Each end of the link uses a signalling-unit error-rate monitor to
|
||
monitor the number of errors at the link level and determine the stability of
|
||
the link. When signalling-link management has determined that the link is
|
||
suitable for use, it will report it to level 3 as being available.
|
||
|
||
Signalling-route management maintains and distributes information and
|
||
distributes information between nodes on the availability of signalling
|
||
routes. Much like a traffic reporter, it sends out messages about the loss or
|
||
degradation of routes causing other nodes to choose alternate routing or take
|
||
appropriate actions.
|
||
|
||
In Fig.6, for example, assume that the link between SSP A and STP 1 failed.
|
||
The STP would send a transfer-restricted (TFR) message to the other SSPs
|
||
informing them that it has limited routing capabilities to access node A. The
|
||
TFR message would contain the point code identifying node A as the subject of
|
||
the message. As long as the other nodes are able to route messages by another
|
||
route, they will not try to access node A through this STP. This helps to
|
||
minimize the traffic between the two STP unless it is absolutely necessary,
|
||
since STP 1 would have to route any messages it received destined for A
|
||
through STP 2. The other network nodes can still route through STP 2 with no
|
||
problem.
|
||
|
||
Since STP 2 will not be able to send messages to SSP A via STP 1 at all, STP 1
|
||
sends a transfer-prohibited (TFP) message to STP 2. This message contains the
|
||
point code for SSP A marking its route as unavailable for messages coming from
|
||
STP 2 in this direction. As you can see the only way STP 1 can get a message
|
||
to SSP A would be to route it through STP 2. It would have to send the message
|
||
right back, causing double traffic over the link joining the two STPs. The TFP
|
||
will prevent that situation.
|
||
|
||
When the route between STP 1 and SSP A is restored, STP 1 will send out
|
||
transfer-allowed (TFA) messages to its adjacent nodes, informing them that
|
||
routing is again available to SSP A. There are additional messages that are
|
||
used to accomplish all the tasks that need to be handled by routeset
|
||
management but this scenario gives you an idea of how nodes communicate the
|
||
availability of routes between each other.
|
||
|
||
The third area of network management is sigalling-traffic management, which is
|
||
responsible for routing the traffic in the network as the availability of
|
||
routes change. Let's take our previous example and look at how traffic
|
||
management handles this situation. At SSP A, all traffic destined for STP 1
|
||
must be stopped and re-routed to STP 2. Link-layer procedures exist to attempt
|
||
to account for all messages which might have been in transit between the nodes
|
||
when the failure occured to ensure that messages are not lost. This
|
||
communication is done using the route through STP 2. This coordination between
|
||
the two nodes terminating the faulty route is called a changeover and is one
|
||
example of how traffic management works in the SS7 network. Traffic which was
|
||
destined for the linkset to STP 1 will now be changed over to the linkset for
|
||
STP 2.
|
||
|
||
Again, there are a number of such procedures that make up signalling-traffic
|
||
management. Congestion procedures were not even mentioned. But network
|
||
management is a big subject - its hard to predict the future, especially with
|
||
the rate of change that's taking place in communications today. However, as
|
||
you read, a great deal of developement is being done in the area of
|
||
centralized services such as Advanced Intelligent Network (AIN). These
|
||
services rely heavily upon the SS7 protocol to communicate.
|
||
|
||
|
||
Glossary of Telephone Network Terms
|
||
|
||
AIN - Advanced Intelligent Network. A network concept in which services are
|
||
created and managed in a centralized location. This moves the service
|
||
intelligence from the telephone office to a service control point.
|
||
|
||
ANSI - American National Standards Institute. Refines the Global SS7 standards
|
||
specified by the ITU-TS for North American and regional Bell operating
|
||
companies.
|
||
|
||
Associated mode - Signalling mode in which a node is directly connected to the
|
||
destination node by a linkset.
|
||
|
||
CLASS - Custom Local Area Signalling Services. A set of services usually
|
||
targeted for residential and small business which provides the equivalent of
|
||
many business features such as caller identification and automatic recall.
|
||
|
||
Connection-oriented signalling - Signalling used to set up, monitor and take
|
||
down calls or pass information related to a call connection.
|
||
|
||
Connectionless signalling - Signalling used to transfer information not
|
||
associated to a particular connection. Often referred to as a Query/Response
|
||
method.
|
||
|
||
FISU - Fill-in signalling units. An SS7 packet sent when there are no MSUs to
|
||
be sent. Since SS7 links transmit a continuous stream of packets, these are
|
||
used as filler when there are no messages which need to be sent.
|
||
|
||
GTT - Global Title Translations. The process of converting digits to an SS7
|
||
address. SS7 uses point codes and subsystems to deliver messages.
|
||
|
||
ISDN - Integrated Services Digital Network. A network concept which provides
|
||
multiple integrated services from a single point of access. ISDN provides
|
||
access to voice, circuit-switched data and packet-switched data as well as
|
||
enhanced call control signalling from the end user to the telephone office.
|
||
|
||
ISUP - ISDN user part. Part of the SS7 protocol which provides connection-
|
||
oriented signalling used for setting up, monitoring and taking down trunks.
|
||
|
||
ITU-TS - International Telecommunications Union-Telecommunications
|
||
Standardization (Sector). Organization that Global SS7 Standards.
|
||
|
||
Link - A communication channel between two adjacent signalling points which
|
||
provides a path for messages to travel.
|
||
|
||
Linkset - A set of links between two adjacent signalling points.
|
||
|
||
LSSU - Link-Status Signalling Unit. An SS7 packet used to convey changes in
|
||
the link state between nodes.
|
||
|
||
MSU - Message-Signalling Unit. An SS7 packet used to send information across
|
||
the network.
|
||
|
||
MTP - Message-Transfer Part. Levels one through three of the SS7 protocol. MTP
|
||
provides reliable transfer of signalling units between network nodes. Its
|
||
responsibilities include point code routing and network management.
|
||
|
||
NSP - Network-Service Part. Refers to the combined services of MTP and SCCP.
|
||
Together, these provide end-to-end application routing.
|
||
|
||
OSI - Open System Interconnection. The telephone hook-up system commonly used
|
||
throughout the world.
|
||
|
||
Point Code - An address for an SS7 network node.
|
||
|
||
Quasi-Associated Mode - Signalling mode in which a message must travel over
|
||
two or more linksets to reach its destination. It is not directly connected to
|
||
the destination point.
|
||
|
||
Route - A path from a signalling point to a destination.
|
||
|
||
Routeset - A collection of routes used to access a destination.
|
||
|
||
SCCP - Signalling-Connection Control Part. Part of the SS7 protocol which
|
||
provides additional routing capabilities to the MTP, including subsystem
|
||
routing and global title translations.
|
||
|
||
SCP - Service Control Point. A database used to access information about calls
|
||
such as routing, billing and the selection of the service provider. The SCP
|
||
provides a centralized form of intelligence for handling calls.
|
||
|
||
SP - Signalling Point. A signalling point that can originate and terminate SS7
|
||
messages but does not have TCAP capability. The term signalling point is
|
||
sometimes used to refer to any network node with signalling capability;
|
||
however this should not be confused with the specific "Signalling Point" node
|
||
type.
|
||
|
||
SSP - Service-Switching Point. A node that can originate and terminate
|
||
messages but does not have the capability to transfer them. It also has the
|
||
ability to send TCAP messages.
|
||
|
||
SS7 - Signalling System 7. A system that specifies the signalling protocol for
|
||
the telephone network.
|
||
|
||
STP - Signalling-Transfer Point. A node used to transfer messages between
|
||
other switching nodes. Acts as a message switching center.
|
||
|
||
subsystem - An application at a node which uses the routing capabilities of
|
||
SCCP.
|
||
|
||
TCAP - Transcaction Capabilities Part. Part of the SS7 protocol which provides
|
||
a generic format for transferring applications-related information.
|
||
|
||
trunk - Facility which carries voice or data traffic between two telephone
|
||
offices.
|
||
|
||
|
||
Information:
|
||
|
||
Signalling System 7
|
||
Travis Russell
|
||
McGraw-Hill
|
||
|
||
Computer Telephony
|
||
Editorial/Business office
|
||
12 West 21 Street
|
||
New York, NY 10010
|
||
tel: 212 691 8215
|
||
fax: 212 691 1191
|
||
Subscriptions
|
||
(free to qualified requesters)
|
||
tel: 800 677 3435
|
||
tel: 215 355 2886
|
||
fax: 215 355 1068
|
||
|
||
|
||
Vendors:
|
||
|
||
Telesoft Design, Inc.
|
||
3475 Lenox Road NE Suite 400
|
||
Atlanta, GA 30326, USA
|
||
tel: 404 238 0528
|
||
fax: 404 235 0529
|
||
email: tsdusa@mindspring.com
|
||
|
||
Telesoft Design, Ltd.
|
||
Unit 1 Luccombe Business Park
|
||
Milton Abbas, Dorset, DT11 0BD, UK
|
||
tel: 44 0 1258 880358
|
||
fax: 44 0 1258 880206
|
||
email: telesoft@tsdesign.zynet.co.uk
|
||
|
||
DataKinetics Limited
|
||
Fordingbridge Hampsire England
|
||
tel: 44 0 1425 655050
|
||
fax: 44 0 1425 655075 |