White
Paper on IP Signalling by IP Alarms Limited (UK)
Author: Steve Nutt
Alarm Installers Guide to IP Signalling - Part 2
We
often hear that one of the great advantages of IP over PSTN is the increased
speed of alarm signal transmission. Whilst this is partly true, alarm
installers need to understand the different stages of an alarm signal
transaction in order to take full advantage of this available speed.
Typically, from the time an alarm panel is activated, it takes approximately
10-15 seconds to dial the telephone number and send a DTMF or FSK signal
over a PSTN line.
When an IP dialler capture device is used, it listens for the telephone
digits, ignores them, plays a handshake, takes the signal, plays a kiss-off
and then transmits the signal over IP. Even though the actual transmission
over IP might take only a couple of hundred milliseconds, the whole process
takes approximately the same time as it does over PSTN. So where is the
speed advantage in using IP?
Well, using the above as an example, it has been wasted waiting for the
panel to go off hook, waiting for the telephone line to settle and waiting
for the telephone number to be dialled - even though it never gets used.
Programming a short or empty telephone number is one way to improve on
this situation, however, installers should also take advantage of any
programmable outputs or relays on the panel where available.
Some IP devices have digital inputs that can be connected to alarm panel
outputs (I/O). These panel outputs change state immediately upon activation
of an alarm and in turn trigger the IP device inputs so that a signal
is sent to the server without delay. Until such time that new alarm protocols
are developed to take full advantage of IP and built directly into panels,
using digital I/O offers the best solution for the retrofit market on
systems where speed of transmission is important.
One area where the speed of transmission is critical is where alarm signalling
is used to trigger the recording of images from IP cameras.
Using IP Signalling to integrate IP Cameras with Alarm Systems
Sophisticated CCTV cameras and transmission equipment were in use long
before many of us had even heard of the Internet. Manufacturers were quick
to update their products to take full advantage of IP, however, the price
tag of such systems remains well out of the reach of the vast majority
of both residential and small commercial Customers. So how can the alarm
industry provide video verification services to this group of Customers
at a more realistic cost?
The answer lies in the use of IP cameras. For alarm installers, their
arrival is akin to the introduction of PIR detectors more than two decades
ago and it won't be long before it becomes routine to fit one in each
protected area as one might do now with a PIR.
So what are our options for integrating IP cameras so that they become
an integral part of a security alarm system?
Most cameras come with some form of built-in motion detection, however,
the unpredictability of this technology prevents it from being used to
trigger an alarm panel into calling a Monitoring Centre. Using motion
detection would also require that there be some way to arm and disarm
the cameras so that they do not send images when people are legitimately
moving around the protected premises. Due to the familiarity and simplicity
of arming and disarming an alarm panel, one option is to have the panel
trigger cameras only when it is armed. This way, images are only sent
to the server in the event of an alarm activation.
This can be done by connecting a cable from the panel to the trigger input
of the camera if it has one. The preferred method however, is to trigger
and control the camera via software commands from the Monitoring Centre
server. The more expensive IP cameras provide a full range of software
commands that allow control of pan, tilt, zoom, pre and post alarm triggering
and other common features. Whilst lower cost cameras do not generally
offer these features, the majority of them do support jpeg snapshots at
various resolutions which provides the industry with a starting point
for mainstream video verification.
Overview of Vendor Neutral Video Verification
The snapshot feature allows an installer to fit a camera from manufacturer
X alongside one from manufacturer Y and Z. This provides great flexibility
for the installer as they are not tied to a specific camera manufacturer.
It is also ideal for the Monitoring Centre as they are not reliant on
a specific server application.
In preparation for viewing and recording images from the protected premises,
the Monitoring Centre database stores various camera related parameters
alongside well known alarm parameters such as event definitions and zone
lists.
Basic camera parameters determine what type of alarm events and which
zone numbers should trigger the recording of individual cameras. They
also specify the total number of images to be recorded and the time interval
between them.
Each IP camera requires a slightly different software command to trigger
it, so the Monitoring Centre maintains a database of custom commands which
they can build up over time to support any brand of camera used by their
installers.
Every second counts in the triggering and recording of images from IP
cameras. Remembering how long it takes to send an alarm signal using traditional
protocols, starting a recording 10-15 seconds after an alarm event would
almost certainly mean losing some of the most important footage.
There are different ways to overcome this depending on the available features
of the alarm panel, the IP alarm device and the camera. A high spec camera
would use a ‘pre-alarm’ feature where it might send the server
ten seconds of footage prior to the alarm and twenty seconds after it.
Lower spec cameras would provide a snapshot command facility which can
be triggered by the use of panel outputs and IP device inputs as already
described.
The Snapshot Command in Action
As an example, we will say that Mr. Smith has an alarm system, an IP alarm
transmitter and an IP camera installed at his home. The transmitter and
camera are plugged into his router and connected via the internet to an
IP enabled Monitoring Centre.
His account number at the Monitoring Centre is 1234 and he has requested
that 10 snapshot images be recorded at 2 second intervals in the event
of an alarm. This information should be entered into the database at the
Monitoring Centre and checked every time an alarm signal is received.
Here are the database fields and data that might be used for Mr. Smith:
Account Number: 1234
Number of Snapshots: 10
Snapshot Interval: 2
Zones That Trigger: 2,3,4
Camera IP and Port: http://mrsmith.dyndns.org:800
Snapshot Command: /CamImage.jpg?UserName=abc&PassWord=xyz
When an alarm signal from account code 1234 is received, the server software
checks the database to find that Mr. Smith has a camera and that images
should be recorded on the server if the alarm was from zones 2,3 or 4
(zones 5,6,7 and 8 might be used to trigger a second camera). The software
concatenates the IP address of the camera and the snapshot command together
so that the following command is sent in a loop, ten times, to Mr. Smith's
camera at two second intervals...
http://mrsmith.dyndns.org:800/CamImage.jpg?UserName=abc&PassWord=xyz
The end result is that ten jpeg images are written onto the hard drive
of the monitoring server and optionally displayed to an operator for further
analysis.
Now let's break down the snapshot command and consider the prerequisites
of Mr. Smith's network and camera setup in order for it to work correctly.
Firstly, you will notice that http://mrsmith.dyndns.org is used instead
of an IP address. This tells us that Mr. Smith does not have a fixed IP
address allocated to him by his internet service provider and that he
has registered something called a dynamic domain name. Put simply, this
is a method of keeping a domain name linked to a changing IP address.
Typically, when a user connects to the Internet, the user's ISP assigns
an unused IP address from a pool of IP addresses, and this address is
used only for the duration of that specific connection. A dynamic DNS
service provider uses a special programme that runs on the user's computer
(some IP cameras also support this), contacting the DNS service each time
the IP address provided by the ISP changes and subsequently updating the
DNS database to reflect the change. In this way, even though a domain
name's IP address will change often, we do not have to know the changed
IP address in order to connect with Mr. Smith's camera.
The :800 in the command lets us know that Mr. Smith's router has been
setup with port forwarding and that port 800 points to his camera. There
is an IP alarm device that can overcome the requirement for port forwarding
in order to trigger cameras, but generally speaking there is no way to
get to Mr. Smith's camera without port forwarding.
Each camera manufacturer has their own set of software commands for controlling
cameras and the /CamImage.jpg? part of our example snapshot command is
just one such variation.
Most snapshot commands contain a question mark and this tells the camera
that a number of parameters will follow. These are normally used to set
the resolution and compression of the image and to provide security credentials.
In summary, with the help of specialist software on the server at the
Monitoring Centre and an IP alarm adapter at the protected premises, low
cost IP cameras can be easily integrated with alarm systems to provide
basic video verification.
‘Any Panel, Any Receiver, Any Monitoring Centre’
Much to the dismay of Alarm installers and Monitoring Centres alike, the
luxury of ‘any Panel, any Receiver and any Monitoring Centre’
has been taken away by the manufacturers of IP alarm adapters and receivers.
For many years, installers have been able to pick their favourite panel
or communicator off the shelf at the distributors, install and programme
it with a choice of protocols, connect it into a telephone line and have
it monitored by any number of accommodating Monitoring Companies. Things
are not nearly as flexible in the move to IP.
Each IP receiver manufacturer has designed their server equipment in such
a way that it will only accept signals from their particular brand of
panel or IP adapter at the client end. So, if you want to upgrade an alarm
system using an IP adapter from a particular manufacturer, you will have
to use the services of a Monitoring Centre that has installed the company’s
brand of IP receiver.
Whilst this may be a little inconvenient for the installer, spare a thought
for Monitoring Centres who are faced with purchasing IP receivers from
each and every one of the manufacturers for fear of losing their subscribers
to competing Monitoring Centres.
As well as doing their best to ensure that their panels only get connected
to their own brand of IP receiver, many manufacturers also offer a ‘universal’
device that will allow any brand of panel to communicate with their receiver.
Unfortunately, these devices are only ‘universal’ at the client
end as they are still locked to a particular brand of IP receiver.
There is a definite requirement for the development of a truly universal
IP alarm protocol so that the industry can return to the much fairer system
of any panel, any receiver, any Monitoring Centre.
In the meantime, there are a small number of universal IP devices on the
market that will communicate with any analogue alarm receiver over IP
to protect the Monitoring Centre's investment in existing equipment.
Web: http://www.ipalarms.net
