How to implement a 100-channel network camera monitoring solution?

1. Calculate line bandwidth

First, we need to determine the bandwidth required for the video output of each camera. The camera parameter introduction generally calls it "compression output bit rate".

For almost all cameras, this parameter can be adjusted manually, and the adjustable range is generally from a few hundred K to 8M or 16M. The adjustable range varies from manufacturer to manufacturer and model to model.

This needs to be decided according to your requirements for image clarity.

For example, here we assume that the compressed output bit rate of each camera is 4 megabytes. For 100 cameras, it is 400 megabytes.

Generally, several cameras share one switch. The switch directly connected to the camera can be a 100M switch. The number of cameras connected to one switch is determined based on the geographical distribution of the cameras.

This type of switch can be selected as a 5-port, 8-port, or 12-port 100M switch, which can connect up to 4, 7, and 11 cameras respectively.

For the sake of convenience, the switch directly connected to the camera can be called the "access layer switch"

Of course, if some cameras are concentrated, a 12-port or 24-port switch can connect 11 or 23 cameras, and the camera output bit rate is relatively large, and the total output bit rate exceeds 100 megabits, then this switch must be a gigabit switch, which can be a 22-port 100M + 2-port gigabit switch.

Multiple access layer switches are connected to a dedicated switch, which is called a "core layer switch". Here, it is necessary to determine whether the aggregation layer switch uses a 100M or 1000M switch, and how many interfaces to use.

If the access layer switch is 8-port, and each is connected to 7 cameras, then 15 access layer switches are needed for 100 cameras. Then the core layer switch can be a 24-port switch, which has 20 or 22 100M interfaces and 4 or 2 Gigabit interfaces.

However, if the access layer switches are gigabit switches, then all ports on the core layer switches must be gigabit.

In a special case, 100 cameras are distributed over a large area and the length of the network cable between the devices cannot exceed 100 meters.

If some access layer switches are far away from the core switches, you can add a transition switch between the access layer switches and the core layer switches to serve as a relay.

This will extend the distance.

If the distance is farther, such as 700 to 800 meters, or even farther, then do not use a switch for relay. Use a pair of fiber optic transceivers, one connected to the core layer switch with a network cable, and the other connected to the access layer switch with a network cable. The two fiber optic transceivers are connected with optical fiber.

However, you need to pay attention to the issue of single-mode fiber and multi-mode fiber. If it is single-mode fiber, the transceiver must also be single-mode. Only one fiber is needed. The distance within a few kilometers does not need to be considered at all. Because there are many technical standards, there are many distances. If it is multi-mode fiber, the transceiver must be multi-mode fiber. A pair of fibers can cover a distance of five or six hundred meters at the shortest and two kilometers at the longest.

In short, no matter how you design it, the actual amount of data flowing through each line cannot exceed the theoretical maximum amount of this line.

2. Calculate the amount of hard disk storage required

After calculation, if the camera compression output bit rate is 1M (note that its unit is b/s or bps, lowercase b means bit), one camera will generate 10GB (note that uppercase means byte) of data in one day.

100 cameras for 30 days would generate 30,000GB of data (note that the compression output bit rate is 1M).

If it is 2M, then the data volume is 60000GB.

If it is 4M, then the data volume is 120000GB.

Based on a 4T hard drive on the market, it will show about 3700GB on the computer. Considering that the file system in the hard drive itself also takes up a certain proportion of space to maintain data, it is estimated that about 3500GB can actually be used.

100 cameras stored for one month on hard disk

1M bit rate, 30000GB, 9 hard drives are required.

2M bit rate, 60000GB, 18 hard drives are required.

4M bit rate, 120000GB, 35 hard drives are required.

8M bit rate, 240000GB, 69 hard drives are required.

Generally, a DVR can only hold one or two hard disks, but some can hold 8 or 12 hard disks. However, the price is not cheap.

You can also use a storage server. 2M bitrate and 4M bitrate require two servers. 8M bitrate can use one 48-disk slot server and one 24-disk slot server.

Or adopt the storage server + hard disk cabinet solution.

You can calculate the prices and weigh the trade-offs of these options yourself.

Take 4 32-channel recorders, each connected to 25 cameras, as an example, calculate how much storage space is needed for 25 cameras. How many hard drives are enough?

If 1 yuan is enough, then there is no problem, the choice is very wide;

If two hard disks are needed, the recorder must support dual hard disks (and also pay attention to the maximum supported hard disk capacity);

If 4 hard disks are needed, the recorder must support 4 hard disks;

If 8 hard disks are needed, then the recorder must support 8 hard disks; (There are only a few machines that support 8 hard disks)

There is also a special case. If your storage capacity is large and 8 hard drives are not enough, you can only reduce the number of cameras connected to each recorder.

For example, if there are 5 video recorders, each connected to 20 cameras, the calculation method is the same as above.

Or 6 recorders, each connected to 16 or 17 cameras, and so on.

Multiple recorders can be connected to small-sized monitors, or share a monitor and switch with a VGA switch, mainly for the convenience of playback or video extraction. Or do not connect to a monitor.

3. Splicing screen

Then you need to decide whether to splice 4 screens (2×2) or 9 screens (2×2) or something else.

Naturally, you need several LCD displays for splicing.

The digital monitoring matrix will have multiple video output signals, which are connected to the LCD display via VGA or HDMI lines.

The digital monitoring matrix is ​​connected to the core switch via a network cable.

This core switch is also connected to the network video recorder.

Logically you can imagine a digital surveillance matrix as a VCR, except that it only displays the video and does not save it.

As for the quantity, generally there is only one digital monitoring matrix, and there are 4 or 9 splicing screens, depending on how many screens you need for splicing.

Of course, choosing a digital monitoring matrix requires how many screens it supports, as well as specific functions.

For example, whether multiple screens can display independently, whether multiple screens can display a picture as a whole, etc.

But don't forget the TV wall mounting brackets. Otherwise, how can you fix 4, 9 or 16 monitors?

Please note that some manufacturers' digital monitoring matrices can install corresponding digital monitoring matrix management software on the computer, which means that you can modify which images the digital monitoring matrix needs to display on the computer.

4. The fourth question

It is indeed as you said, 22 access switches enter the computer room with the help of fiber optic transceivers and connect to the core switch.

The core switch is connected to the NVR. But if you choose a 24-port switch, there are only 2 ports left to connect to the NVR.

That means you can only connect 2 NVRs.

Unless you choose a 36-port switch, but the cheapest 36-port Gigabit switch costs seven or eight thousand.

You can use some alternative methods at will

Workaround 1

The entire monitoring network is divided into two parts, using two 16-port Gigabit switches to connect to 11 access switches respectively.

In this way, there are 5 ports left, 1 port is connected to the top switch, and the remaining 4 ports can be used to connect to the NVR. A maximum of 4 ports can be connected.

Here we assume that your network will use 4 NVRs, so only 2 NVRs need to be connected to one core switch.

The same goes for the core switch on the other side.

A 16-port Gigabit switch is cheaper.

Two 16-port Gigabit switches are connected with another 8-port, 12-port or 16-port Gigabit switch. The specific number of ports to use depends on how many ports the TV wall device (digital monitoring matrix) requires.

Workaround 2

There is a separate gigabit switch next to the core switch, which is specifically used to connect multiple NVRs.

In this way, only 22+1 ports are used, and 1 port is left.

By the same token, if you want to use equipment such as a TV wall, you can connect a gigabit switch next to the core switch to use up all the network ports of the core switch.

The digital surveillance matrix is ​​connected to this dedicated Gigabit switch.

Workaround 3

In the actual wiring process, some access switches (A) can be connected to another access switch (B). The access switch (A) is not directly connected to the core switch, but the access switch (B) is connected to the core switch, so that the access switch (A) is indirectly connected to the core switch through the access switch (B).

This method can consider the connection method based on the actual geographical distribution of your cameras.

Just ensure that the actual required bandwidth of the network cable between the access switch (B) and the core switch is less than 100 Mbps.

That is to say, the number of cameras connected to the access switch (B) should include the number of all cameras under the access switch (A).

This method is used when the two access switches are close to each other and can be connected with a network cable.

This also saves an optical fiber and a pair of optical fiber transceivers.

Of course, considering that there are multiple NVRs (assuming 4) and digital surveillance matrices (assuming 4 network cables), many network ports are needed. A total of 8 ports are needed, so only 16 ports of a 24-port switch can be used to connect to the switch.

Of the 22 access switches, only 16 can be directly connected to the core switch, and the remaining 6 access switches are connected to another access switch.

Of course, these three methods can be used in combination.

However, it should be noted that no matter how the network is connected, the maximum bandwidth of each network cable must not exceed the maximum value of the network cable itself (it is recommended to discount the theoretical maximum by 10%, 100 Mbps as 90 Mbps, 1000 Mbps as 900 Mbps, just to be on the safe side)

The digital monitoring matrix can also be connected in the following ways:

There will be many network ports on a single digital monitoring matrix, just connect all of these network ports to the core switch.

The digital monitoring matrix will also have multiple VGA or HMDI interfaces, which can be used to connect to multiple LCD monitors separately.