Video over IP

Bandwidth Considerations for IP Surveillance Cameras & How to reduce the bandwidth occupied by them

While there are a lot of advantages of IP Surveillance over Analog Surveillance, bandwidth occupied by IP Surveillance Cameras are an important consideration before implementing an IP surveillance project. In this article, let us look at the factors affecting the bandwidth occupied by each video stream of IP Surveillance Cameras and certain measures that can help reduce the bandwidth occupied by them.

Factors affecting the bandwidth occupied by IP Surveillance Cameras:

Resolution: Various resolutions are supported by IP Surveillance cameras, from QVGA to HD. Higher the resolution, higher the bandwidth required to transmit every video stream from the IP Camera to the user monitoring station (computer/ server).

Frame Rate: Various frame rates are supported by IP Surveillance cameras from 5 frames per second to 30 frames per second (for example). Higher the frame rate, higher the bandwidth required.

Compression codec: There are various types of compression schemes applied to the transmitted video like MPEG4, H.264, etc. Better compression (H.264, for example), lower the bandwidth required to transmit each video stream.

Compression ratio: A video stream can be compressed more than 90% of its original size, based on the compression codec and the compression ratio used. Better the compression ratio, lower the bandwidth required.

Type of Camera Operation: An IP Camera can be operated in fixed mode or PTZ mode. In the latter, the movement of the camera can controlled (even from a remote location) and the bandwidth occupied in this mode is higher.

Type of movement/complexity in the scene: If the movement/ complexity in a scene being shot by the IP Camera is less, the bandwidth occupied by the resultant video stream is lesser.

Other factors: Many other factors like indoor/ outdoor shooting, shooting in bright light/ dark surroundings, etc affect the amount of bandwidth occupied by each video stream. Special applications like face recognition/ license plate recognition require higher resolution/ frame rate recording and hence they will occupy more bandwidth.

Let us look at the approximate bandwidth occupied by two types of IP Cameras (VGA, 1.3 Mega pixel) at various frame rates, below.

Type: VGA Resolution

5 frames per second – 256 Kbps to 750 Kbps
8 frames per second – 384 Kbps to 1 Mbps
10 frames per second – 500 Kbps to 1.2 Mbps
15 frames per second – 750 Kbps to 1.2 Mbps
30 frames per second – 1 Mbps to 1.5 Mbps

Type: 1.3 Mega pixel

5 frames per second – 750 Kbps to 1.5 Mbps
8 frames per second – 1.2 Mbps to 2.5 Mbps
10 frames per second – 1.5 Mbps to 3 Mbps
15 frames per second – 2.5 Mbps to 4 Mbps
30 frames per second – 2.5 Mbps to 5 Mbps


These values are approximate and are provided here just to get a rough idea. The actual bandwidth occupied by each video stream varies based on a lot of factors, as mentioned above. Each IP Camera requires bandwidth as mentioned above, and for a whole IP Surveillance project one needs to add up all the bandwidth occupied by the video streams generated by every IP Camera used in the project.

Though Local Area Networks (LAN) can offer 1 Gbps of bandwidth today (and can go up to 10 Gbps), the bandwidth is shared with a number of other devices/ applications occupying the network. Multiple users accessing the same video stream (simultaneously) and multiple users accessing various video streams from remote locations need to be considered, in addition to the number of video cameras being monitored over the LAN.

So, it is very important to optimize the bandwidth occupied by the IP Surveillance cameras.

How to reduce/ Optimize the bandwidth occupied by IP Surveillance Cameras?

Some techniques that can be followed to reduce/ optimize the bandwidth occupied by IP Surveillance cameras are given below.

During nights / weekends, generally no one is expected to be in the campus being monitored. So, we can set the IP surveillance cameras to start recording (and transmitting) video streams only when some kind of motion is detected. That is the part which is required to be monitored anyway.

When multiple people need to view the same video stream (from a particular IP camera), it is better if the stream is sent as a multicast. Multicast sends one stream to multiple monitoring/ viewing stations (for the image output from the same camera) as opposed to unicast which sends one video stream per each monitoring/ viewing station. Both the IP Cameras and the network equipments should support multicast, to take advantage of this technique.

Certain IP Cameras enable streaming at constant bit rates. This ensures that there is no bandwidth spike if sudden excessive motion/ detail is observed in the scene. But the video quality is obviously reduced during those times. A camera that supports variable bit rate occupies more bandwidth during high detail/ fast motion recording and the network infrastructure should be ready to accommodate for those bandwidth spikes.

If an IP Camera is set to higher compression ratios, it is possible to achieve lower bandwidth. But the IP Camera should have adequate processing capacity and it should support such advanced compression techniques.

Setting a higher frame rate for local viewing (LAN) and lower frame rate for remote viewing (WAN) can reduce the bandwidth required to see the video streams from a remote location. Obviously, the image quality is compromised but various devices (like cell phone screens, etc) that are used to see these remote video streams support a lower resolution anyway.

You could stay up to date on the various computer networking/ related IT technologies by subscribing to this blog with your email address in the sidebar box that says, ‘Get email updates when new articles are published’


  • perminus


    I’ve a quick one kindly assist,after setting up my camera’s using point to point with airgrid m5.
    The issue i’ve is that the images are not smooth, showing in slow motion modes with flips thought.
    Kindly advice the way forward to resolve this.
    Camera are 2MP.