What Cabling and Network Infrastructure Does a Commercial Security Camera System Need?

Choosing cameras is usually where a commercial security camera project starts — resolution, field of view, low-light performance, brand. But cameras are only one half of a working surveillance system. The other half is the network behind them: the cabling that reaches each camera, the power that runs to it, the switch it plugs into, the uplink that carries its video, and the recorder or platform that stores and manages it. Skipping that planning doesn't mean it goes away — it just gets discovered later, usually as a change order or a camera that never quite performs the way it was supposed to.

This article is about that infrastructure: what's behind the camera, not the camera itself. For camera features, monitoring, access control, and the rest of a security system, see our commercial security page. For the cabling and network side of a project, WOW Electronics designs and installs structured cabling and business network infrastructure for commercial clients across the Yuma area, and that includes the network a camera system runs on.

The Basic Path Behind an IP Camera System

Many current commercial surveillance systems use IP cameras that communicate over network infrastructure, rather than a separate coaxial system. In general terms, the path a camera's video takes looks something like this: the camera connects through horizontal network cabling to a switch (often a Power over Ethernet, or PoE, switch, or a recorder with built-in PoE ports); that switch connects — directly or through an uplink/backbone — to wherever the video is recorded and managed, whether that's a network video recorder (NVR), a video management system (VMS) server, a cloud-connected gateway, or some combination of these.

That's a useful mental model, but it isn't the only way a camera system can be built — the right topology depends on the site, the camera count, the building layout, and the recording platform chosen (more on those options below), so no single diagram fits every project. What matters for planning purposes is that every link in that chain — cable, power, switch, uplink, recording platform — has to be sized and located correctly for the cameras it's supporting.

Cabling to Each Camera Location

A conventional IP-camera design commonly runs an individual structured-cabling connection to each camera location, similar to how a data drop is planned for a workstation or an access point. That's not a universal technical requirement for every possible camera or system architecture, but it's the starting assumption worth planning around for a typical commercial multi-camera install.

For each planned camera location, the practical planning questions are:

• Where exactly will the camera be mounted, and what pathway will the cable actually take to reach it?
• Is the run entirely indoors, or does it cross an exterior wall, an unconditioned space, an underground path, or an exposed outdoor pathway?
• What cable jacket, conduit, or pathway protection is appropriate for that specific environment?
• Where does the run terminate — back to a nearby switch, a recorder, or a telecommunications room?
• Which switch or recorder port is intended to serve that camera?
• Is there a reasonable expectation of future cameras being added nearby, and does the pathway leave room for that?

Because each wired camera generally needs its own connection back to a switch and rarely moves once mounted, this cabling is sometimes run as a direct-connect termination rather than through a wall outlet — our guide to planning office data drops covers that distinction and how many connections a project actually needs. And because camera cabling is easiest to plan before walls and ceilings close, our article on prewiring before drywall goes up covers why that timing matters, including for camera systems specifically.

Cat6, Cat6A, or Fiber for Camera Runs

The deeper comparison between Cat6, Cat6A, and fiber is its own topic — see our comparison of Cat6, Cat6A, and fiber for that full discussion. For camera-specific planning, a few points are worth knowing:

Cat6 is commonly adequate for many individual camera runs. Cat6A may be selected where a project's broader performance goals, PoE requirements, environment, or future-network plans justify it — not because any particular resolution automatically demands it. Camera resolution alone doesn't determine which cable category a project needs; the decision depends on the whole design, not just the camera. Fiber is commonly used for backbone or uplink connections — between telecommunications rooms, over longer distances, or between buildings — rather than for individual camera-to-switch runs, though the right choice always depends on the specific project.

No single cable category is correct for every camera project, and none guarantees future compatibility on its own — that depends on the whole system it's connected to.

How Far Can Camera Cabling Run?

Standard balanced twisted-pair Ethernet cabling has a design distance limit, and this project has consistently sourced that limit as 90 meters for the permanent link, with the complete channel — the permanent link plus the patch cords at each end — capped at 100 meters overall, per ANSI/TIA-568.2-E (corroborated by independent manufacturer-technical sourcing, including Fluke Networks). The 100 meters is the total channel figure, not an additional 100 meters on top of the 90; that's the standards-based ceiling for a compliant copper horizontal run, camera cabling included.

If a planned camera location is farther from the nearest switch or telecommunications room than a standard copper run can reach, that's usually a sign the design needs another distribution point closer to the camera, a different switch location, a fiber backbone to a nearer distribution point, or another appropriate media choice — not that the camera can't be served at all. It's also worth checking whether the specific switch, recorder, or camera has its own manufacturer-stated distance limitations beyond the standards-based figure, since those are product-specific and can vary.

Power Over Ethernet: Two Separate Questions

PoE lets a single cable carry both data and power to a camera, which can eliminate the need for a separate electrical run to every mounting location. But planning a PoE camera system means answering two separate questions.

Can that specific port deliver the power that specific camera needs? IEEE 802.3 defines PoE in types and classes, and the type available at a given switch port sets a ceiling on how much power that port can deliver. Under the current 802.3bt framework, Type 1 delivers up to roughly 15 W at the source with about 13 W available at the device; Type 2 delivers up to roughly 30 W at the source with about 25.5 W available at the device; Type 3 delivers up to roughly 60 W at the source with about 51 W available at the device; and Type 4 delivers up to roughly 90 W at the source with about 71 W available at the device, once cable loss is accounted for. A fixed camera with basic IR illumination may need very little power. A pan-tilt-zoom (PTZ) camera with motors, heaters, multiple sensors, or strong illuminators can need considerably more — sometimes enough to require Type 3 or Type 4 PoE specifically. Matching the camera's actual maximum power draw, from its manufacturer specification sheet, to what that port can deliver is a real design step, not a formality.

Can the switch deliver that much power to every connected camera at the same time? This is the question "24 PoE ports = 24 cameras" skips. A switch's total PoE power budget is a shared resource across all of its active ports. Depending on the switch, that total budget may be lower than what would be required to deliver every port's maximum supported PoE output simultaneously. A switch can have plenty of ports and still run out of power headroom before it runs out of ports, particularly once higher-draw cameras — PTZs, cameras with heaters for outdoor use, or cameras with strong IR arrays — are part of the mix. Sizing a PoE switch means totaling the manufacturer-stated maximum power requirement of everything that will be connected to it, not assuming a typical or average draw, and comparing that total against the switch's actual power budget with reasonable headroom for the system to operate reliably.

Choosing a Switch for a Camera System

Port count is one input into switch selection, not the whole answer. Relevant questions can include how many camera ports are actually needed now and reasonably soon; the PoE type and total power budget each port and the switch need to support; how much uplink capacity the switch needs toward the recorder or backbone; where it will physically live and what rack space is available; any redundancy requirements; and what management or monitoring the system calls for.

No single switch brand or model is correct for every project — the right switch is the one sized to the actual camera load, power budget, and uplink requirements, and not every system needs a fully managed switch.

Bandwidth and Uplink Planning

It's tempting to size a network by port speed alone — "these are all Gigabit ports, so we're fine" — but port speed and actual traffic are different things. A 1 Gbps camera port describes the capacity of the Ethernet link; it does not mean the camera continuously generates 1 Gbps of video traffic, and twenty camera ports running at Gigabit speed doesn't mean twenty cameras' worth of Gigabit traffic is actually moving through the network at once.

What actually matters is aggregate bandwidth — the combined, real traffic every connected camera generates, and where that traffic has to travel: from each camera to its local switch, from that switch onward to wherever video is recorded, and between distribution rooms or buildings if the design has more than one. Camera bandwidth itself isn't a fixed number tied to resolution. According to Axis Communications' own technical guidance on network video bitrate, actual bandwidth depends on the combination of resolution, compression method and level, frame rate, scene complexity, and motion in the scene — a camera pointed at a quiet hallway and an identical camera pointed at a busy loading dock can produce meaningfully different bitrates even at the same resolution and frame rate. Camera count alone does not determine network bandwidth. The reliable way to size a system is to look at the actual cameras being selected and their planned recording settings — tools such as AXIS Site Designer estimate bandwidth and storage from the actual camera models and scenario settings (camera model, scene detail, lighting, motion, recording type, number of streams, frame rate, resolution, codec, and compression) rather than a flat per-camera number.

It's also worth being clear about where that traffic goes. In an on-premises architecture, camera video can remain on the local network as it moves between cameras, switches, and the recording platform. Cloud-connected or hybrid designs may send some video or related traffic off-site, depending on the platform — that's a function of the specific platform chosen, not something true of every camera system by default.

Where the Video Is Recorded and Managed

The recording and management platform changes what the network around it needs to look like. Common approaches include a dedicated NVR with built-in PoE switch ports for a self-contained system; standalone PoE switches feeding a separate VMS server; a fully server-based VMS on general IT infrastructure; cloud-managed or cloud-recording platforms where some or all video is sent off-site; and edge-recording where cameras handle some local storage themselves. None of these is universally correct — the right one depends on the project's scale, IT environment, and how the client wants to access footage.

What matters for infrastructure planning is that the choice affects where traffic flows, how the uplink and switch get sized, where the recording hardware physically lives, whether WAN/internet bandwidth needs planning, and how storage gets planned. Choosing the recording platform is part of the network design conversation, not a decision made after the cabling is already in the walls.

Storage and Retention: What the Network Connection Affects

Storage and bandwidth are related, because most of the same variables that drive bandwidth — resolution, compression, frame rate, scene activity, and continuous versus event-based recording — also drive how much video is generated per unit of time, which is what accumulates into storage. Retention duration works differently: it doesn't affect bandwidth, but it directly multiplies how much total storage that video adds up to over the length of time it's kept. There's no reliable universal figure for storage per camera or per retention period; those numbers shift with every one of those variables. AXIS Site Designer, for example, estimates storage from the selected camera and scenario settings together with the planned retention period, which is a more reliable starting point than a flat rule of thumb.

Multiple Telecommunications Rooms and Larger Sites

On a larger site, cameras are sometimes served from more than one telecommunications room rather than one central location, particularly once camera clusters are far enough apart that running every camera back to a single switch isn't practical within standard cabling-distance limits. That can mean PoE switches located closer to groups of cameras, with backbone connectivity — which may use fiber depending on the site and distance — carrying traffic back to a central distribution point or recording platform. Our article on MDF and IDF planning covers how and when a site needs more than one equipment room in more depth; a camera system is simply one more thing that planning needs to account for.

Cameras Across Multiple Buildings

When camera coverage spans more than one building — a main building and a detached warehouse, a gate house, a parking structure — the connection between those buildings becomes part of the design question, alongside everything else that needs to travel between them. Our comparison of wireless bridges and fiber between buildings covers that decision in more depth; a handful of cameras on a detached structure is one of several factors that can go into it, not a decision made in isolation from the rest of the building's connectivity needs.

Outdoor and Perimeter Cameras

For Yuma commercial properties, exterior and perimeter camera locations — parking areas, loading docks, fence lines, building exteriors — can introduce planning issues that interior runs may not. Exterior runs need pathway and cable construction appropriate to the environment they're actually installed in: heat and UV exposure, weather, and whether the run is exposed, buried, or protected in conduit. Some perimeter runs may be longer than interior runs, building-to-pole transitions need their own planning, pole-mounted cameras raise their own access and serviceability questions, and any outdoor electrical or surge-protection work should follow the applicable standards and the equipment manufacturer's own requirements.

No single exterior cable type or pathway approach is correct for every outdoor camera — the right answer depends on the specific environment, run length, and mounting situation, which is why exterior runs are worth designing individually.

Upgrading From Existing Analog or Coax Systems

A common question for a retrofit project is whether existing coaxial cabling — left over from an older analog camera system — can be reused when upgrading to IP cameras. The honest answer: sometimes, but it has to be evaluated, not assumed either way. Existing coax doesn't automatically mean a full recable is required, and it also doesn't mean the existing cable should automatically be reused. Products exist that carry network data and PoE over coaxial cable for exactly this kind of migration — Axis Communications, for example, publishes PoE-over-coax adapters with stated distance and power capabilities that vary by the coax gauge in the wall and the camera's power draw. Whether that approach makes sense depends on the condition of the existing cable, the distances involved, and the power the chosen cameras need — a real option worth evaluating on a retrofit, not a default recommendation for every one.

Camera Traffic and Network Segmentation

Camera traffic is often planned as its own logical segment of the network, or otherwise intentionally managed as part of the overall network design, rather than mixed in with general office traffic without any planning. This isn't just a general networking habit — Axis Communications' AXIS OS Hardening Guide recommends isolating Axis devices and related infrastructure from the production/business network, using mechanisms that can include VLAN segmentation, limited routing, firewalling, or access control lists, depending on the environment. Exactly how that gets implemented is a network-design and IT decision specific to that business's infrastructure — worth a conversation between whoever is deploying the cameras and whoever manages the network, so the two are designed together rather than colliding after the fact.

Can You Use Your Existing Business Network?

A common assumption is: "We already have switches and Ethernet running through the building, so can't we just plug the cameras into what's already there?" Sometimes — but it needs to be checked, not assumed either way. It's the same evaluation this article has already walked through — available ports, PoE type and total power budget, uplink capacity, cabling reach to each camera location, and rack space — applied to what's already installed instead of a clean design. An existing business network might be entirely adequate for a camera system, or it might need targeted upgrades in one or two of those areas. The only way to know is to check the actual existing capacity against the actual planned cameras, rather than assuming either outcome.

Planning for Future Camera Expansion

Camera systems may grow over time as coverage needs change, buildings are added, or new areas need monitoring. Worth thinking about during initial design: unused rack space and switch port capacity, capacity margin on the backbone and uplink, pathways left for additional camera runs, and whether future buildings or areas are part of the client's plans. How much headroom makes sense depends on the client's actual growth plans and budget, not a fixed formula.

A Hypothetical Example: Working Through the Decisions

To show how these pieces fit together, consider a hypothetical commercial property with a main office/warehouse building and a detached storage building, planning a mix of interior cameras, exterior perimeter cameras, and a few cameras on the detached building.

The process starts with camera locations, not port counts: where does the client want coverage, and what does each location need to reach it? Interior cameras get planned as individual cabling runs back to a nearby switch or telecommunications room, following the same pathway/distance/termination questions as any other data drop. Exterior and perimeter cameras get evaluated for their own environment rather than assumed to match the interior runs. The cameras on the detached building raise the building-to-building question, resolved with the same wireless-bridge-versus-fiber evaluation any other connection between those buildings would use — not decided in isolation just because it's carrying camera traffic.

From there, the design works out from the cameras to the switch — what PoE type does each selected camera need, and does the planned switch have enough total power budget for all of them running at once, including any higher-draw PTZ or illuminated exterior units. Then out to the uplink — what realistic aggregate traffic will these specific cameras generate, and does the backbone connecting the buildings and reaching the recording platform have real capacity for that traffic plus everything else already using it. Finally, the recording platform gets chosen, which feeds back into where equipment lives and how much uplink and storage capacity the design needs.

Nothing here implies a fixed bandwidth-per-camera number or a single correct topology — it's meant to show the order of the decisions, not to represent every project's design.

What to Plan Before Cameras Go In

Planning the Network Before the Cameras Go In

A security camera system is only as good as the network carrying its video, and that network is easiest to get right when it's planned alongside the cameras — not discovered as a problem after installation. WOW Electronics is licensed in Arizona under ROC #330717 and designs and installs both commercial security camera systems and the structured cabling and network infrastructure behind them for commercial clients across the Yuma area — cabling, PoE and switch capacity, backbone connectivity, and coordination with whatever recording platform the project calls for. If you're planning a multi-camera system and want the network side worked out alongside the camera selection, that's a conversation worth having before either one is finalized.