4G Connectivity for Deployable CCTV Towers

Why connectivity on a CCTV tower is harder than it looks

A deployable CCTV tower – the kind deployed on construction sites, temporary events, retail parks, and highways – looks like a straightforward connectivity problem. Put a 4G router in the box, give it a SIM, done. In practice, every aspect of that environment creates constraints that a standard office or vehicle router was never designed to handle.

The tower is solar and battery powered, often operating for days or weeks without any direct sunlight. The cellular connectivity unit draws from the same limited energy budget as the cameras themselves. The mast is exposed – rain, temperature extremes, vibration from wind loading. The site may be in a rural or semi-urban location with marginal cellular coverage from a single operator. And the entire system may be unattended for months at a time, with no engineer on call to swap a frozen router or failed SIM.

These constraints mean that a product designed specifically for this environment will outperform a repurposed office router in almost every measurable way – power consumption, reliability, remote management capability, and long-term operating cost. This guide covers the full architecture of what that product needs to look like.

The power budget: solar, batteries, and consumption reality

A deployable CCTV tower runs from a photovoltaic panel and a battery bank – typically 12V or 24V lead-acid or LiFePO4. The solar charge controller (most commonly a Victron SmartSolar MPPT in the UK market) manages charging and load management. The entire electrical system operates from the battery rail, and every component draws from a limited daily energy budget that shrinks to near-zero during extended overcast periods.

Understanding the power budget is the starting point for any connectivity design. A tower that runs four cameras, a connectivity unit, and associated electronics in the UK winter – where available solar energy can drop to 10-15% of peak summer capacity for days at a stretch – needs every component to earn its place on the load list.

Typical power budget for a 4-camera tower

The sleep/wake architecture

The correct architecture for a solar-powered CCTV tower separates the always-on control layer from the power-hungry cellular radio. A low-power embedded computer (a Raspberry Pi 5 draws around 3-5W at idle, significantly less in a suspend state) manages all local functions – reading the Victron MPPT via VE.Direct, monitoring digital inputs, controlling camera power via relays, and logging GPS position. The cellular modem is powered up only when there is data to transmit or a remote management session is active.

This two-layer architecture – always-on control at low power, modem sleeping between transmission events – can reduce daily connectivity energy consumption from 360Wh to 80-120Wh. On a tower with a 200Ah battery at 24V (4,800Wh total), that difference represents days of additional autonomy during a UK winter.

Most off-the-shelf 4G routers have no mechanism for this. They run continuously, drawing 8-15W regardless of whether any data is being transmitted. A purpose-built connectivity unit with GPIO-controlled modem power is a meaningful design advantage.

Powering cameras: PoE, passive PoE, and the 12V rail

Cameras on a CCTV tower are almost universally powered over Ethernet – but there is an important distinction between IEEE 802.3af/at/bt PoE (which negotiates power delivery via a protocol handshake at 48V) and passive PoE (which injects a fixed voltage onto the cable without any negotiation).

Standard PoE switches are designed for 48V DC systems. A solar-battery tower operates from a 12V or 24V rail. Bridging that gap requires either a DC-DC boost converter to generate 48V for a proper PoE switch (adding cost, complexity, and inefficiency), or the use of passive PoE at 12V directly from the battery rail.

Passive PoE at 12V: why towers use it

Most IP cameras rated for outdoor or vandal-resistant use accept a wide input voltage range – commonly 12V DC or PoE 802.3af – because the camera’s internal PSU regulates to whatever the camera electronics need. In practice, most cameras happily accept 12V passive PoE injected onto the spare pairs of a Cat-6 cable (pins 4/5 positive, 7/8 negative in the standard passive arrangement).

This allows the tower to distribute camera power from its 12V or 24V battery rail directly, without any intermediate conversion. Each camera port on the connectivity unit gets a passive PoE injector, a relay to individually switch power, and a current sensor to detect fault conditions (a camera drawing zero current indicates failure; one drawing abnormally high current may indicate a short).

Individual port switching via relay

One feature that separates a purpose-built tower connectivity unit from a standard PoE switch is per-port power switching. A relay on each camera port allows the control system to cycle power to a specific camera remotely – without affecting the other cameras, the router, or the rest of the system. In practice, a frozen or crashed camera that is not responding to network commands can be hard-rebooted from thousands of miles away in seconds.

Standard PoE switches have no remote per-port power cycling capability outside of SNMP-managed enterprise switches that were never designed for 12V rail operation. This is a concrete functional advantage of a purpose-built product.

System architecture overview

Why Cat-6 is the standard in UK CCTV tower deployments

Experienced UK CCTV tower manufacturers specify Cat-6 as standard for all internal and external cabling runs. This is not just a case of specifying the newer standard for its own sake – there are concrete technical reasons why Cat-6 outperforms Cat-5e in this specific application.

Cross-talk and PoE at distance

Cat-6 cable has significantly tighter pair twisting and thicker insulation than Cat-5e, which dramatically reduces alien cross-talk (interference between pairs from adjacent cables bundled together in a conduit). In passive PoE applications – where 12V is injected on the spare pairs and the data signal runs simultaneously on the other pairs – the physical separation of pairs in Cat-6 reduces the risk of the PoE voltage inducing noise onto the data pairs. At cable runs of 20-50m (typical for a CCTV tower with a run from the electrical cabinet to the camera housing at the top of the mast), this matters.

Higher power capacity for future PoE standards

Cat-5e is rated for carrying up to 60W per pair under 802.3bt (PoE++). Cat-6 handles the same load with lower resistance, which means less voltage drop and less heat generated in the cable jacket. In an IP65 enclosure in direct sunlight, heat management is not trivial. Specifying Cat-6 now also future-proofs against cameras that demand more power – thermal imagers and PTZ cameras with heater elements routinely draw 25-30W, which stresses Cat-5e at any meaningful cable length.

Robustness of outdoor-rated Cat-6 jackets

Outdoor Cat-6 cable – direct burial or UV-stabilised PVC jacketed – is physically more robust than equivalent Cat-5e in most product lines. The heavier gauge of Cat-6 conductors (typically 23 AWG vs 24 AWG in Cat-5e) is more resistant to damage during installation in conduit. For a product that is installed by different teams with varying levels of cable handling care, the physical durability margin matters.

RF environment on a CCTV mast

A CCTV tower with a 5G MIMO antenna array at the top is generating radio frequency emissions in close proximity to the cabling runs. Cat-6’s superior shielding characteristics – and the availability of F/UTP and S/FTP variants with foil shielding – provide meaningful additional isolation compared to Cat-5e in this RF environment. Unshielded Cat-5e running near SMA-connected 5G antenna cables can act as a receiving antenna for interference. The performance difference may not be significant in most installations, but for a manufacturer whose tower must perform consistently in marginal conditions, the margin is worth having.

The connectivity unit: what it needs to do

Current deployable CCTV towers typically assemble their connectivity from separate components: a 4G router, a PoE switch, a GPS tracker, a Victron display or separate monitoring device, and whatever digital IO capability the manufacturer has chosen to add. Each component has its own power supply, its own management interface, and its own failure mode. The total power draw is the sum of all these devices running simultaneously.

The case for a purpose-built connectivity unit is that a single device, designed specifically for this environment, can replace all of them – at lower total power consumption, with a unified management interface, and with integration that the separate-components approach can never deliver.

The minimum viable feature set

Cellular technology: why LTE Cat 6 and 5G are the right choices

The cellular category specification of the modem is the single most consequential technical choice in the connectivity unit. Get it wrong and the tower will fail to connect on sites where coverage is marginal – exactly the sites that need it most.

Why LTE Cat 4 is not enough

LTE Category 4 modems – which were the standard in most 4G routers until around 2020 – offer theoretical downlink of 150 Mbps but, critically, are limited to a single downlink carrier. In rural and semi-rural UK deployments, where coverage often comes from a single tower with congested spectrum, a Cat 4 modem cannot perform carrier aggregation. It is stuck on one band at a time. When that band is congested, performance drops. When there is no signal on the primary band, there is no fallback within the LTE layer.

LTE Cat 6 and carrier aggregation

Category 6 modems support 2x carrier aggregation (2x CA) in downlink – they can bond two LTE carriers simultaneously, doubling effective throughput and, more importantly, increasing the number of cells and spectrum blocks the modem can access. In a marginal coverage environment, this makes a material difference to reliability. A Cat 6 modem that can combine a weak signal on 800 MHz with a stronger signal on 1800 MHz will maintain a better connection than a Cat 4 modem limited to the weaker of the two. Theoretical peak for Cat 6 is 300 Mbps downlink, though real-world performance in a rural tower installation is typically 20-80 Mbps.

5G Sub-6 and why it matters now

5G NR Sub-6 GHz deployment in the UK has progressed rapidly since 2022. EE, Vodafone, O2, and Three have all extended their 5G footprints to major towns, arterial roads, and an increasing proportion of the semi-rural locations where CCTV towers are deployed. A 5G NR modem offers not only higher throughput but also improved spectrum flexibility – and 5G NR’s beamforming capabilities at 3.5 GHz can maintain a usable connection to a distant 5G NR cell where a 4G modem on the same band would have dropped out.

The Teltonika Calyx 5G HAT+ is one example of a 5G NR modem designed for integration into embedded systems – it provides 5G Sub-6 with 4x MIMO, GPIO-controlled power and reset, and a wide operating temperature range (-40 to +75 degrees Celsius) suitable for outdoor enclosures in UK conditions. It interfaces to a Raspberry Pi via the 40-pin GPIO header, which keeps the integration clean and allows the Pi to manage modem power directly.

Relevant UK band support by LTE category

IoT SIM cards for CCTV towers

The SIM card is the most overlooked component in a CCTV tower connectivity design. Most manufacturers specify a single-network SIM from a consumer or business mobile contract, which works until the tower is deployed to a site where that operator has marginal coverage. The result is a support call, a site visit, and a SIM swap – at significant cost relative to the price of a better SIM in the first place.

Multi-network IoT SIMs

A multi-network or roaming IoT SIM connects to whichever of the available UK networks provides the strongest signal at the deployment site. Unlike a locked consumer SIM, a properly configured multi-network IoT SIM will roam between EE, Vodafone, O2, and Three domestically, selecting the best available network automatically. For a device that may be deployed anywhere from a city centre construction site to a rural highways project, this is not a luxury – it is a requirement.

The distinction between a genuine multi-network SIM and a “preferred network” SIM is important. Some IoT SIM products have a preferred home network with fallback to others only when the primary network has no signal – this is better than a single-network SIM but still falls short of true multi-network behaviour where all networks are treated as equal options. For CCTV tower deployments, specify a SIM with genuine domestic multi-network operation.

Private APN and static IP

Many tower deployments benefit from a private APN – a dedicated access point name that routes traffic through the SIM provider’s network without touching the public internet. This has two practical advantages for CCTV towers. First, the device can be assigned a static IP address, making remote access straightforward without requiring a VPN or dynamic DNS workaround. Second, traffic between the tower and the control centre travels on a private network segment, reducing exposure to public internet threats.

For deployments where video is being streamed to a Video Management System (VMS) hosted in a private data centre or on a corporate network, a private APN is the cleanest architectural solution. The tower connects to the APN, traffic routes directly to the private network, and the VMS sees the camera streams as if they were on the local LAN. See IoT SIM Explained for a detailed treatment of APN types and private connectivity options.

Data consumption planning for CCTV towers

The data consumption implications of continuous cloud streaming are significant enough to make local NVR storage the preferred architecture for most CCTV tower deployments. The tower records locally; the connectivity link is used for management, alerting, and on-demand access – not continuous video exfiltration. This reduces data costs by an order of magnitude and reduces the criticality of the cellular link from “must be up 100% of the time” to “must be up when needed.”

Antenna selection and placement on a CCTV mast

The cellular modem is only as good as the antenna system feeding it. A 5G NR modem connected to a poorly specified antenna at the wrong location on the mast can perform worse than a Cat 6 modem on a correctly specified antenna. In a marginal coverage environment, antenna design is not secondary to modem choice – it is equally critical.

MIMO and why 4×4 MIMO matters for 5G

Modern 5G NR modems support 4×4 MIMO (Multiple Input, Multiple Output) – four independent transmit and receive antenna chains that allow the modem to spatially multiplex data across the same frequency bandwidth. This provides both a throughput benefit (in strong signal conditions, 4×4 MIMO can approximately double throughput versus 2×2 MIMO) and a resilience benefit – if one antenna path is degraded, the others continue to function.

The practical implication for CCTV tower design is that the connectivity unit needs four separate antenna ports, and the installation needs four antenna elements with adequate spacing between them. The spacing requirement for effective MIMO is a minimum of half a wavelength at the operating frequency – at 3.5 GHz (5G NR n78 band), this is approximately 43mm. At 800 MHz (Band 20), it is approximately 187mm. Most CCTV mast installations have sufficient physical space to meet this requirement.

Antenna types for mast installation

Placement on the mast

Antenna placement on a CCTV mast involves two competing constraints. Higher placement gives better line of sight to distant cell towers – the mast itself provides the elevation. But the CCTV cameras and their housings are also at height, and their metal structures can cause reflections and nulls in the antenna pattern. For a mast installation, the antenna elements should be placed above the camera housings where possible, or on the side of the mast most favourable to the direction of the serving cell tower.

The five SMA bulkhead connectors on the connectivity unit enclosure (four for the 5G modem, one for GPS) should be mounted on the top face of the enclosure where pigtail runs to external antennas are shortest. The GPS antenna benefits particularly from clear sky view – a magnetic base GPS antenna mounted on top of the enclosure with unobstructed sky view above will achieve lock much faster than one inside or below the mast structure.

Remote access, VPN, and GPS

WireGuard over OpenVPN for battery-constrained devices

VPN protocol selection matters more than it might appear in a power-constrained deployment. OpenVPN, the traditional choice for remote access on industrial routers, is CPU-intensive – it runs in userspace and requires significant processing per packet. WireGuard, the modern alternative supported by most Linux kernels since 5.6, runs in kernel space and is dramatically more efficient. On a Raspberry Pi or similar ARM-based compute board, the CPU load of maintaining an active WireGuard tunnel is a fraction of the equivalent OpenVPN connection. For a device that is trying to minimise power consumption during periods of low activity, this matters.

WireGuard also has faster handshake and reconnection behaviour, which is beneficial in the CCTV tower use case where the cellular link may drop temporarily (cell reselection, network congestion, brief signal loss) and needs to re-establish quickly when connectivity returns. The correct architecture is for the WireGuard tunnel to be established only when the modem is powered up, and torn down cleanly when the modem sleeps – this avoids keepalive traffic consuming data when no active session is needed.

GPS for asset protection

Tower theft is a real and recurring problem in the UK. A tower with GPS position logging can trigger an immediate alert if the unit is moved more than a defined radius from its last known position. Combined with tilt detection from an IMU (which identifies the moment the mast is lowered for theft), the system can send an alert before the tower has been removed from site – providing the response window that human intervention requires.

The u-blox NEO-M8N module is a widely used GPS/GNSS receiver for this application – it achieves 1-2m position accuracy, supports GPS, GLONASS, Galileo, and BeiDou, and draws under 30mA active. At 100mW on a system drawing 30-40W, GPS adds negligible power consumption while providing a theft deterrent capability that adds meaningful value to the product proposition.

The ideal product: a wishlist for manufacturers

No product currently on the UK market fully addresses all of these requirements in a single unit. The closest approaches assemble most of them from separate components at the cost of higher total power draw, more complex installation, and multiple management interfaces. The gap between what exists and what would be ideal is clear, and it represents a genuine product opportunity.

eSIM, eUICC, and SGP.32: the remote provisioning case

Every tower that uses a physical SIM carries an inherent operational risk: if the operator providing that SIM loses coverage at the deployment site – due to a cell outage, spectrum refarming, or simply a change in the competitive landscape of UK MVNO pricing – the only remedy is a physical site visit to swap the SIM. For a tower on a six-month highway project two hours from the nearest depot, that is a meaningful cost.

eUICC and remote profile management

An embedded Universal Integrated Circuit Card (eUICC) is a SIM that holds multiple operator profiles and can switch between them under software control. The traditional M2M eUICC specification (GSMA SGP.02) allows profiles to be downloaded and switched remotely via an SM-DP+ (Subscription Manager Data Preparation) server, operated by the SIM provider. If operator A’s coverage degrades, the platform can push a profile switch command to the device, which activates operator B’s profile without any physical intervention.

For CCTV tower deployments, this transforms a potential site visit into a remote management action. The tower stays connected; the SIM platform handles the operator switch in the background. The requirement for this is that the connectivity unit’s SIM slot accepts an eUICC (MFF2 soldered format for outdoor environments, or a standard SIM slot accepting an eUICC form factor card), and that the cellular module firmware supports GSMA remote SIM provisioning commands.

SGP.32: IoT-specific eSIM management

GSMA SGP.32, ratified in 2023, is the IoT-specific eSIM architecture that removes some of the complexity of the M2M SGP.02 stack. It introduces a simplified eIM (eSIM IoT Manager) function that does not require the device to be connected at the moment of profile provisioning – the profile can be staged and downloaded opportunistically when the device next has connectivity. This is directly relevant to CCTV towers, which may have intermittent connectivity during low-battery states or network maintenance windows.

SGP.32 also simplifies the bootstrap profile mechanism – the device needs only a basic connectivity profile to pull down a full operational profile, rather than requiring the full M2M provisioning infrastructure from day one. For a product being deployed to dozens of sites with different operator preferences, the operational flexibility this provides is substantial.

For a deeper treatment of eUICC architecture and SGP.02 vs SGP.32, see euicc.co.uk.

5G RedCap: the future of low-power CCTV connectivity

Current 5G NR modems – including the Teltonika Calyx HAT+ referenced in the prototype BOM – are full-feature 5G implementations designed for maximum throughput. They support 4×4 MIMO, 100 MHz channel bandwidth, and high peak data rates. All of that capability comes at a cost: power consumption that is higher than a comparable LTE Cat 6 modem, and a hardware cost premium that may be difficult to justify for applications that never need gigabit throughput.

A CCTV tower streaming four 1080p H.265 camera feeds uses at most 8-12 Mbps of uplink and rarely more than a few Mbps of downlink for management traffic. The full 5G NR feature set is architectural overkill for this application – and the modem’s power budget reflects that mismatch.

What RedCap changes

3GPP Release 17 introduced Reduced Capability (RedCap) NR – a new device class designed to bridge the gap between power-hungry full 5G NR and the low-throughput NB-IoT/LTE-M category. RedCap offers downlink rates up to 220 Mbps (more than enough for any CCTV tower application), uplink up to 100 Mbps, and dramatically lower device complexity and power consumption than full 5G NR – achieved by reducing MIMO layers to 2×2, limiting channel bandwidth to 20 MHz, and removing features like full carrier aggregation that are irrelevant for IoT applications.

The benefit for a solar-battery CCTV tower is concrete: a RedCap modem consumes meaningfully less power than a full 5G NR modem while maintaining 5G network access, 5G latency characteristics, and 5G NR coverage as operators continue to expand. It is 5G connectivity at a power budget closer to LTE Cat 6.

UK operator deployment timeline

As of early 2026, UK operators are in varying stages of RedCap deployment. The technology requires software updates to existing 5G NR base stations rather than new hardware at the cell level – operators with mature 5G infrastructure are better positioned to enable RedCap quickly. EE and Vodafone have both signalled RedCap support in their 5G roadmaps, with commercial availability expected to expand through 2026 and 2027. RedCap devices launched today are likely to find growing network support within their operational life on deployed CCTV towers.

For detailed coverage of RedCap’s UK status and the device ecosystem developing around it, see 5gredcap.co.uk.

Power consumption comparison: standard vs purpose-built

The following diagram illustrates the difference in daily energy consumption between a CCTV tower assembled from separate components and one using a purpose-built connectivity unit with sleep/wake management.

Frequently asked questions