Telematics & Connectivity for Connected Vehicles
← Back to: Automotive Electronics Assemblies
This page is written to help you turn “we need connectivity” into a concrete telematics ECU plan: which radios to use, how to partition the system, what to ask for on security, antennas and power, and which BOM fields really matter. The goal is that you can brief suppliers and internal teams with clear, actionable requirements instead of vague wishes.
Telematics / Connectivity – Role & Scenarios
A telematics ECU is the connectivity hub of a modern vehicle, combining cellular, Wi-Fi, Bluetooth, GNSS and on-board security. It bridges in-vehicle domains and cloud services, enabling OTA updates, remote diagnostics, fleet monitoring and app-based connected features while keeping data flows controlled, observable and traceable.
Early vehicles often carried a single cellular box dedicated to eCall. It only woke up during an accident, sent a small data burst and supported a voice channel to emergency services. Modern platforms instead use a telematics ECU as a multi-radio connectivity hub that stays online for most of the vehicle’s lifetime.
This hub concentrates cellular, Wi-Fi, Bluetooth and GNSS, then exposes secure links to the gateway, ADAS and perception domain controllers, the infotainment head unit and sometimes a separate V2X module. Those domains own the driving and safety decisions; the telematics ECU owns the secure, reliable pipe that connects them to backend services.
In practice, the same box that pulls OTA images from the cloud also uploads diagnostic logs, streams fleet usage metrics and maintains app-based features such as remote lock, HVAC pre-conditioning and charge scheduling. The following sections translate these scenarios into concrete data flows and bandwidth requirements.
Connectivity Use Cases & Data Flows
Telematics connectivity workloads fall into a few repeatable patterns. Grouping them makes it easier to estimate uplink and downlink demand, set latency targets and decide where dual-SIM or backup links are justified. This section stays at the use-case and data-flow level, without going into protocol-stack details.
Safety & Regulatory – eCall, eDR, Emergency Services
Safety services rely on the telematics ECU to place an emergency call, send the vehicle’s position and provide a snapshot of critical parameters after a crash. Data volume is modest but call setup time and availability are tightly constrained by regulation and OEM safety goals.
Uplink: small bursts of positioning and crash data plus voice. Downlink: control commands and operator interaction. Availability: very high; many platforms use backup power and prefer primary networks or dual-SIM strategies for this traffic.
Data & OTA – Firmware Updates and Telemetry Upload
Firmware and software updates pull large images from the cloud into the vehicle, often during parked windows. At the same time, diagnostic logs and telemetry are pushed back in batches or on events. These flows dominate long-term data consumption on many fleets.
Uplink: periodic or event-driven log bundles; bandwidth depends on logging density and compression strategy. Downlink: high-volume OTA images with flexible timing but strict integrity and security requirements. Availability: high but not life-critical; retries and scheduling can hide most network outages.
Infotainment & User-Facing – Hotspot and Remote Functions
User-facing connectivity includes Wi-Fi hotspots for passengers, phone projection and app-based remote features such as locking, HVAC pre-conditioning and charge control. These workloads shape the perceived responsiveness and quality of the connected-car experience.
Uplink: mixed traffic from apps and projected devices, typically moderate but bursty. Downlink: streaming and web content with higher bandwidth demand when multiple users are active. Latency: seconds for remote control is acceptable, while interactive apps prefer sub-second round-trip times.
Fleet & Shared Mobility – Operations and Usage-Based Insurance
Fleet and shared-mobility operators depend on frequent position and status updates to run dispatch, billing and maintenance. Usage-based insurance adds fine-grained driving metrics on top, sometimes combined with video or ADAS event summaries.
Uplink: regular telemetry points plus higher-density data for certain vehicles or time windows. Downlink: configuration changes, geofence updates and control messages. Availability: high for operations; long-running fleets often justify robust roaming, eSIM profile management or dual-SIM to avoid coverage gaps across regions.
Radio Architecture Overview (Cellular / Satellite / BT / Wi-Fi / GNSS)
A typical telematics ECU aggregates several radio subsystems behind a single housing. The cellular chain combines a modem SoC and RF transceiver with external PA, LNA, filters and antenna switches. Wi-Fi and Bluetooth usually share a combo chip, while GNSS can be integrated or implemented as a dedicated receiver for positioning and timing.
Cellular Modem Chain
The cellular path starts with a 4G or 5G modem SoC that hosts the protocol stack and, in some designs, runs application workloads. A separate RF transceiver drives front-end components: power amplifiers, LNAs, band filters, duplexers and antenna switches. Diversity and MIMO paths add extra RF chains to improve throughput and reliability.
Wi-Fi / Bluetooth Combo and GNSS Receivers
Wi-Fi and Bluetooth are often integrated into a single combo IC with one or more RF chains (for example, 2×2 MIMO) covering 2.4 GHz and 5/6 GHz bands. External front-end modules may add PA and LNA stages when range or regulatory margins demand it. GNSS capability may reside inside the modem or combo chip, or it may be provided by a dedicated RF plus baseband device supporting GPS, Galileo, BeiDou and GLONASS on L1/L5 bands.
Antenna Sharing and Coexistence
A shark-fin assembly typically hosts several antennas: cellular main and diversity, Wi-Fi/Bluetooth, GNSS and sometimes V2X. Some paths share radiators via switch matrices and filters, while others use dedicated feeds for sensitivity and isolation. Poor coexistence planning quickly shows up as degraded GNSS performance or reduced Wi-Fi throughput when the cellular PA is active.
All-in-One Modules vs Discrete Radio ICs
Many platforms use pre-certified cellular modules that integrate the modem, RF front end and sometimes GNSS, reducing RF design and regulatory workload. High-volume or high-end platforms may prefer discrete modem, RF and combo chips to optimise band sets, MIMO configuration and cost. The trade-off is greater flexibility and reuse at the cost of deeper RF and certification competence.
System Partitioning & Host Interfaces
Inside the telematics ECU, compute resources are commonly split into three domains: the cellular modem domain, which runs the air-interface protocol stack; an application processor that hosts telematics software; and a security domain that holds keys, certificates and the root of trust for secure boot and updates.
CP, AP and Security Domains
The modem or CP domain terminates the 3GPP stack and is tightly coupled to operator certification. The AP MCU or SoC implements diagnostic protocols, OTA orchestration, fleet applications and integration with in-vehicle networks. A dedicated HSM or secure element protects credentials, enforces secure boot and may also support V2X security in some architectures.
In-Vehicle Network Connections
The telematics ECU usually connects to the rest of the vehicle via CAN or CAN FD and, on newer platforms, one or more Automotive Ethernet links such as 100BASE-T1 or 1000BASE-T1. Entry-level designs may only exchange diagnostic data with a gateway, while centralised E/E architectures route high-volume traffic through Ethernet to domain controllers or a central compute ECU.
Host Interfaces to Radio Devices
High-speed data between the host processor and the cellular modem is typically carried over PCIe, USB or SDIO, with UART reserved for control and diagnostics. Wi-Fi combo devices commonly use PCIe or SDIO for data plus UART or PCM for Bluetooth control and audio. GNSS receivers stream NMEA or binary messages via UART, with I2C or SPI used for configuration and, in some cases, timing outputs such as PPS.
Stand-Alone Telematics ECU vs Integrated Connectivity
A stand-alone telematics ECU gives clear functional and safety boundaries and can be reused across vehicle lines, at the cost of extra housing and harness complexity. Integrating connectivity into a head unit or gateway reduces ECU count and may share compute and storage, but couples operator certification and telematics updates more tightly to those larger systems.
Security, eSIM & Credentials Management
A secure telematics ECU combines connectivity with strong identity and credential protection. The security architecture spans eSIM or iSIM for operator access, secure boot and update chains anchored in hardware, and dedicated storage for keys and certificates that back mutual TLS and, in some designs, V2X security domains.
eSIM, iSIM and Multi-Profile Connectivity
Embedded SIMs replace removable cards with a soldered eUICC that speaks ISO 7816 to the modem. Integrated SIM, or iSIM, goes further by placing SIM functions inside the modem or SoC. Both support multiple operator profiles, allowing a single unit to switch coverage across regions, contracts or backup subscriptions when roaming or operator performance changes.
These capabilities depend on a secure element that implements GSMA specifications and a protected interface so that profile downloads, activation and deactivation cannot be spoofed. Multi-SIM designs may expose several SIM channels or combine a discrete eSIM with an iSIM, but always rely on hardened storage for identities and life-cycle state.
Secure Boot and Secure Update
Secure boot starts from a small hardware root of trust, typically ROM code and fused keys, that verifies the next stage loader and then cascades signatures down to the operating system and applications. Hardware engines for AES, SHA and RSA/ECC offload crypto operations so large OTA images and signatures can be processed efficiently, even on constrained telematics SoCs.
Anti-rollback mechanisms prevent reverting to older, vulnerable firmware versions. Monotonic counters, version fields stored in one-time-programmable regions and secure policy logic in the boot chain ensure that only signed images with allowed or higher revision numbers can run, even when attackers gain temporary access to the update channel.
Keys, Certificates and Security Domains
Device keys and certificates should never be stored in plain flash. Instead, a hardware security module, secure element or TPM-like device isolates root keys, long-lived client certificates and sensitive secrets. The telematics stack then uses these anchors to establish mutual TLS with backend services so both the cloud and vehicle can authenticate each other before exchanging data.
Some architectures place V2X certificates and crypto accelerators into a separate security domain, isolating safety-critical broadcast messaging from general telematics functions. In that case, the telematics ECU interfaces with the V2X domain through defined APIs while key management and credential rotation for V2X remain in a dedicated subsystem.
Threat Model and Hardware Countermeasures
Typical threats include stolen modules, abused diagnostic ports, fake base stations and attempts to load unauthorised firmware. Hardware defences include secure boot, debug port lockdown, encrypted storage, tamper-resistant secure elements, and secure erase paths when a device is decommissioned. A clear threat model tied to hardware capabilities helps ensure that connectivity does not become the weakest link in the vehicle.
Performance & Antenna / Power Considerations
Radio performance is shaped by link KPIs, antenna placement and power behaviour. Throughput and latency determine whether OTA, hotspot and remote features feel responsive. GNSS TTFF, sensitivity and handover behaviour impact fleet tracking and safety services. On the hardware side, antenna integration and power strategy set the limits that software and networks must live within.
Connectivity KPIs and Vehicle Use Cases
Uplink throughput drives log uploads, fleet analytics and usage-based insurance, while downlink throughput dominates OTA campaigns and passenger hotspot usage. Latency and jitter matter for remote control, interactive apps and session setup times such as eCall call establishment limits. Handover performance determines how gracefully connections survive highway speeds and border regions.
GNSS adds its own KPIs, including cold-start time-to-first-fix, warm and hot start performance and tracking sensitivity in dense urban conditions. These metrics feed directly into how quickly a parked vehicle can regain a position fix and how robust location and timing data are for telematics and safety workloads.
Antenna Placement and Vehicle Integration
Roof-mounted shark-fin modules offer clear sky view and reduced body shielding for GNSS and cellular, at the cost of tight packaging and styling constraints. Hidden or glass-integrated antennas can improve aesthetics but depend heavily on roof materials, glass composition, ground reference and nearby metal structures, all of which affect gain patterns and efficiency across bands.
Getting antenna decisions right requires bringing RF engineers into early body and trim design reviews. The choice of shark-fin location, available cable routes and acceptable ground plane sizes sets hard limits on what any modem or GNSS receiver can achieve, no matter how capable the silicon appears on paper.
Coexistence and In-Vehicle Noise Sources
Coexistence uses spatial, frequency and time separation to keep radios out of each other’s way. Physical spacing and careful orientation between antennas improve isolation. Filters and duplexers separate nearby bands such as Wi-Fi and some radar or V2X channels. Scheduling can stagger heavy transmissions to avoid worst-case self-interference in shared housings or harnesses.
High-current DC-DC converters, traction inverters and electric power steering introduce broadband noise that couples into harnesses and grounds. TCU placement, grounding strategy and power filter design must consider these sources so that sensitivity and throughput remain acceptable in real driving conditions, not only in a quiet lab.
Power Modes, DRX and 12 V / 48 V Supply Behaviour
Average power draw is dominated by idle and sleep behaviour rather than rare peak transmit bursts. Modem features such as PSM, eDRX and DRX reduce wake frequency at the cost of slower downlink responsiveness. Application processors and Wi-Fi radios contribute their own sleep and wake strategies, triggered by timers, user activity or vehicle bus events.
When the telematics ECU shares 12 V or 48 V rails with other loads, cold crank, load dump and power sequencing must be considered. The design should keep modems and storage within their voltage and inrush limits and define a controlled start-up and shut-down order so that unexpected power events do not corrupt flash or leave the unit in an undefined state.
IC & Module Selection Map (7 Brand + Others)
When you plan a telematics ECU, you are not choosing a single chip—you are assembling a connectivity stack. This map helps you think in blocks: modem or module, RF front-end, Wi-Fi/Bluetooth, GNSS, security and the power tree. For each block you can decide which parameters matter most and which vendors, including the seven major brands, are realistic candidates.
Cellular Modem / SoC
For cellular you first choose the service profile: 4G Cat-4/Cat-6 for mainstream data, Cat-M/NB-IoT for low-power boxes, or 5G NR (SA/NSA) when you need high throughput or long-term headroom. You then map this to band sets, regional coverage, MIMO configuration and automotive grade. A key decision is whether you use a pre-certified module or a bare modem SoC.
In practice, most 4G/5G platforms come from dedicated modem and module vendors. The seven major brands tend to appear around the modem rather than inside it: Ethernet gateways, secure MCUs and power management. You can treat the modem or module as a connectivity engine and use the 7-brand ecosystem to wrap it into your vehicle network and safety concept.
RF Front-End: PA Modules, LNA, Filters and Switches
The RF front-end turns modem I/O into real RF power at the shark-fin. Here you care about band coverage (low/mid/high bands), output power and linearity, receive chain noise figure and the number of MIMO or diversity paths. You also have to fit everything into a small, hot enclosure with the right package and AEC-Q temperature grade.
Many PA modules, LNAs and antenna switches come from specialist RF vendors or are hidden inside modules. You normally lean on those suppliers for front-end design and validation. The seven major brands are stronger in power, protection and interface ICs that sit around the RF front-end rather than inside it.
Wi-Fi / Bluetooth Combo
For Wi-Fi and Bluetooth you start from user-facing features: hotspot performance, phone mirroring and audio. That leads to a decision on Wi-Fi standard (Wi-Fi 5 vs 6/6E), MIMO (1×1 vs 2×2), supported bands and BT 5.x with or without LE Audio. On the hardware side you match host interfaces (PCIe, SDIO, UART, PCM) and antenna count to your main SoC and shark-fin layout.
Automotive-grade Wi-Fi/BT combos are offered by a smaller group of connectivity vendors, sometimes bundled with infotainment or gateway platforms from the seven brands. You can either follow the Wi-Fi/BT solution that comes with a chosen SoC platform, or deliberately pick a separate combo device when you need more control over coexistence and roadmap.
GNSS Receivers
GNSS selection starts with positioning accuracy and timing needs. If you just need basic tracking, a single-band L1 receiver integrated in the modem or module may be enough. If you want faster TTFF, better urban performance or timing features, you look at multi-band receivers that track GPS, Galileo, BeiDou and GLONASS with PPS and time-sync outputs.
Several of the seven brands offer GNSS-related products or tightly integrated platforms, and there is also a strong set of dedicated GNSS specialists. You can keep the decision simple: integrated GNSS for mainstream telematics boxes, or a standalone automotive-grade GNSS IC when timing and robustness are central to your design.
eSIM, Secure Element and HSM
For identity and credentials you choose between a removable SIM, a soldered eSIM or an integrated iSIM. On top of that you decide how much secure storage you need for keys, certificates and counters, and which crypto engines should be available in hardware. Your goal is to anchor secure boot, mutual TLS and profile management in a tamper-resistant hardware domain.
Here several of the seven major brands are direct suppliers: secure elements, HSMs and TPM-like devices with automotive-grade temperature and long-term availability. You can mix integrated security inside the modem or SoC with an external secure element to keep long-lived credentials portable across platform generations.
PMIC and Power Tree
Finally, you turn the vehicle’s 12 V or 48 V supply into a clean power tree for modem, SoC, Wi-Fi/BT, GNSS and security parts. You check VIN range, cold-crank and load-dump limits, the number of rails, sequencing requirements and monitoring functions. Your PMIC choice directly influences thermal behaviour, efficiency and how gracefully the telematics ECU rides through power events.
This is where the seven major brands are strongest. You can either pick platform PMICs tailored to a specific SoC or modem, or build a modular tree from automotive buck/boost converters, load switches and supervisors. The key is to document rail voltages, currents and sequencing rules clearly so power and connectivity teams are working from the same map.
BOM & Procurement Checklist for Telematics
When you send an RFQ for a telematics ECU, you want suppliers to understand exactly what you need without pages of email back-and-forth. This checklist turns the design discussions from this page into BOM fields you can drop straight into your templates. You can keep the wording compact but still capture the decisions that matter for performance, approvals and lifetime.
Cellular Requirements
For the cellular block, your BOM should at least spell out: region band set (target markets and mandatory bands), LTE/NR category (Cat-4, Cat-6, Cat-M, NB-IoT, 5G NR SA/NSA) and MIMO configuration (for example, 2×2 on downlink). You also call out whether you need Cat-M or NB-IoT fallback for low-power or backup coverage.
You then add the approval expectations: PTCRB/GCF and key regional carriers, plus eCall and VoLTE support where required. If you want GNSS integrated in the modem or module, you state that explicitly, together with the automotive grade and temperature range you expect. These few fields set the tone for which modem and module families suppliers can realistically propose.
Wi-Fi / Bluetooth Requirements
For Wi-Fi and Bluetooth you can be equally direct. You specify the Wi-Fi standard (5, 6 or 6E), bands (2.4/5/6 GHz) and antenna count so suppliers know whether you are targeting 1×1 or 2×2 MIMO. You mention coexistence expectations—for example, that the combo must behave well next to LTE and V2X inside a shared shark-fin.
For Bluetooth you confirm the BT 5.x version, whether you rely on LE features or classic hands-free audio and whether LE Audio is on your roadmap. You close this group with the automotive grade and temperature range so suppliers do not answer with consumer-only chipsets.
GNSS Requirements
On the GNSS side, you tell suppliers whether you want a single-band receiver or a multi-band L1/L5 solution, and which constellations are mandatory: GPS, Galileo, BeiDou and GLONASS. You describe sensitivity in application terms—urban canyons, indoor parking, fast start after long park time—instead of quoting abstract dB numbers.
If timing is important, you add requirements for PPS or time-sync outputs. Finally, you state if you accept integrated GNSS inside the modem or Wi-Fi/BT combo, or if you prefer a standalone automotive-grade GNSS IC with its own antenna feed. That choice has big impact on both RF layout and supplier shortlist.
Security and Credentials Requirements
For security you describe what type of subscriber identity you expect: removable SIM, soldered eSIM or integrated iSIM. You add whether remote profile provisioning and multi-profile support are mandatory. You then state your expectations for secure storage size and which crypto engines the platform must offer in hardware: AES, SHA and RSA/ECC at a minimum.
To tie this into your backend, you add the certificate model: device ID, mutual TLS with the cloud and any separation you plan between telematics and V2X security domains. You close with a simple yes/no on secure boot and anti-rollback being mandatory and whether you want a discrete secure element or are willing to rely on the modem or SoC’s integrated HSM.
Power and Supply Requirements
On the power side you specify the VIN range including cold-crank and load-dump behaviour, the peak current budget during transmit and OTA, and an average current target for parked operation. You can phrase this as “vehicle must survive X days parked without draining below Y%” so suppliers can size low-power modes accordingly.
You also state whether you expect the vendor to deliver a complete power tree or only the module’s internal rails. If sequencing matters for modem, SoC and storage, you list that expectation so PMIC proposals include power-good and reset logic. A short note on inrush handling and EMI behaviour will save you time later in design reviews.
Environment and Reliability Requirements
Environment fields anchor the telematics ECU into your vehicle-level spec. You define the temperature grade, vibration and shock environment, and whether the board will see conformal coating or potting. You also mention humidity and corrosion expectations if the ECU is mounted in exposed locations.
Finally, you include a simple lifecycle expectation: for example, a 10–15 year support horizon and typical minimum supply years. This helps you avoid being locked into short-lifecycle consumer silicon for a long-lived vehicle platform.
Supply-Chain and Approval Risks You Should Call Out
On top of pure BOM fields, it is worth writing down a few expectations around supply-chain risk. If you choose the module route, you acknowledge that a single component carries PTCRB, GCF and regional approvals. A later module change can trigger re-certification, so you ask suppliers how long each module is expected to stay active and what their replacement policy looks like.
If you choose the chip route, you gain more control but also take on more RF, power and test responsibility. In your RFQ you can be explicit about which approvals you expect the supplier to cover and which ones you plan to own. The same applies to security devices: swapping secure elements or HSMs later affects your PKI and cloud infrastructure, so you want suppliers to declare lifecycle, drop-in compatible options and migration plans up front.
FAQs – Telematics / Connectivity Planning
These twelve questions condense the telematics and connectivity topic into decisions you can actually use when you scope a new ECU or vehicle platform. I write the answers so you can reuse the text in RFQs, internal checklists or social posts. The visible answers below stay aligned with the FAQ structured data on this page.
1. When do I need a dedicated telematics box instead of integrating connectivity into the head unit?
I move to a dedicated telematics box when connectivity is safety relevant, when it must survive several head unit generations or when I want a separate security and approval domain. If OTA, fleet services and regulations drive my roadmap, keeping a stable TCU and letting infotainment evolve like consumer electronics is usually the cleaner long term choice.
2. How do I choose between 4G, 5G and satellite backup for a new vehicle platform?
I start from coverage and the data model. For mainstream cars with moderate OTA and app traffic, solid 4G still works well. I look at 5G when I expect heavy data, long platform life or features that benefit from low latency. Satellite backup only makes sense if the use case truly leaves normal cellular coverage.
3. What are the key cellular bands and certifications I should check for global vehicle deployment?
For global deployment I build a simple band matrix across my target regions and insist suppliers clearly list which bands are fully supported. On top of that I check PTCRB or GCF, regional approvals and any eCall or emergency requirements. If a module misses one critical band or approval, my rollout plan will feel it later in deployment.
4. How should I size uplink bandwidth for OTA and data logging workloads?
I usually start with a simple budget per car and per day. OTA defines how many gigabytes I need to move in a given time window and logging defines how often I upload events or traces. If a back of the envelope calculation already looks tight, I either push for higher categories or rethink how much data I really transmit.
5. When does it make sense to use a modem module versus a chipset-level design?
A modem module pays off when I need fast time to market, limited RF and certification effort or when volumes are moderate. A chipset level design only makes sense once I have strong RF and test resources and a platform big enough to amortise the extra work through lower BOM cost and tighter integration across several model years.
6. How do I manage antenna placement and coexistence for cellular, Wi-Fi, BT and GNSS in one vehicle?
I bring RF engineers into body design early and treat the shark fin as a shared resource, not a cosmetic part. Cellular and GNSS get the cleanest view, then I place Wi-Fi and Bluetooth with enough spacing and filtering. Time and power coordination between radios helps me avoid worst case self interference inside a cramped housing on the roof.
7. What are best practices for eSIM, profile management and multi-SIM redundancy in telematics units?
I treat eSIM profiles like long lived contracts. Remote provisioning and at least one backup profile per key region give me room to react when coverage or commercial terms change. For demanding fleets I may add multi SIM redundancy, but I always anchor profile operations in a secure element that follows GSMA rules and my backend policies.
8. Which hardware features matter most for secure boot and OTA update of a telematics ECU?
For secure boot I insist on a ROM based root of trust, hardware engines for AES, SHA and public key algorithms, and a protected place to store keys and certificates. I also look for a clear anti rollback mechanism and version tracking so nobody can push my telematics ECU back to older firmware with weaker security controls.
9. How do I protect the vehicle against fake base stations and man-in-the-middle attacks at the hardware level?
I accept that I cannot block every fake base station, but I can make it very hard to impersonate my backend. Hardware secure elements hold device identities, all critical traffic uses mutual TLS and I lock down debug and update paths. Clear logging and detection hooks then let higher layers spot suspicious behaviour and react quickly.
10. What GNSS features are essential for fleet tracking and usage-based insurance?
For fleets and usage based insurance I care most about fast time to first fix after long parking, robust fixes in urban canyons and reliable timestamps. Multi constellation support is almost mandatory now. I add multi band and clean timing outputs when I want better accuracy and easier correlation between in vehicle events and backend records.
11. How should I connect the telematics unit to the in-vehicle network—CAN, Ethernet or both?
I treat CAN or CAN FD as the baseline for diagnostics and safety related signals. When I plan heavy data upload, rich services or a zonal architecture, I add at least one automotive Ethernet link and terminate it at a gateway or central compute node. On complex platforms a mix of CAN and Ethernet gives me both flexibility and clean isolation.
12. What are typical power consumption targets for always-connected telematics units in parked vehicles?
I do not chase a magic milliamp number. Instead I decide how many days the vehicle should survive without charging and how much battery capacity I can spend on telematics. From there I back calculate an average current target and use it as a hard budget that modem, Wi-Fi and application teams have to design around.