Capacity

Capacity planning determines how much wireless throughput the deployment must support and how many cells are required to meet that demand.

Why Capacity Comes First?

Capacity is the foundation of the network design. Getting the capacity requirement wrong can lead to an inappropriate technology or architecture choice resulting in an under- or over-sized deployment.

Because Onyx LTE and 5G use distinct hardware platforms, technology selection should be considered early in the design. Coverage, Architecture, and Accessories then build on the capacity requirement established here.

This is why the ROM AP Calculator asks for the application-level inputs early in the workflow and why its Technology Assessment should be considered alongside the calculated capacity requirement.

Do this in the Calculator

1

Set Capacity & Application Inputs

In the Capacity & Application panel, set:

• Technology: 4G / LTE or 5G NR
• Channel BW: (Only for 5G NR) 20 / 40 / 100 MHz
• Primary Use Case: The category that best represents the dominant application (General Enterprise / Voice & Data, IoT/Telemetry & Sensors, Video Surveillance & Broadcast, Logistics/AGVs & Automated Vehicles, Machine Vision & Real-time Control, AR/VR & High-bandwidth Immersive)
• Latency Requirement: Not Sensitive / Moderate / <30 ms (LTE delivers >50 ms; 5G NR delivers <30 ms)
• Downlink / User: <3 / 3–50 / >50 Mbps
• Uplink / User: <1 / 1–10 / >10 Mbps
• Concurrent Devices: <25 / 25–100 / 100+
• Device Mobility: Fixed / Mobile / Continuous

Figure: ROM Calculator - Capacity & Application Panel

The calculator uses these inputs to estimate traffic demand and determine the capacity-driven AP count.

2

Review Estimated APs

In the Estimated APs panel, review:

• Capacity estimate: n APs (DL needs X · UL needs Y): The estimated AP count and the calculated downlink and uplink requirements. The larger of X and Y is your capacity-driven cell count, and the calculator identifies the binding direction, such as "UL-limited".

Figure: ROM Calculator - Estimated APs Panel

• Technology Assessment: Indicates whether the selected technology is a good fit, with the reasoning behind the recommendation.

Figure: ROM Calculator - Technology Assessment Panel

Understanding the Methodology

The Calculator performs the capacity calculation automatically. This section explains the methodology so that you can validate the recommendation and adapt it when the deployment does not fit the Calculator's generic application categories.

The Calculator's built-in capacity estimate applies a 0.65 concurrency factor to the device count in your selected Concurrent Devices bucket before computing throughput demand: it never evaluates 100% of devices as simultaneously active.

The manual method below (Application Profile → Total Throughput Demand) does not apply this factor, because it uses your own per-application peak-concurrency figures directly: a number you've already defined as "simultaneously active during peak operating conditions." This is intentional as the manual method is more precise when you know your actual peak concurrency, while the Calculator's 0.65 factor is a generic estimate for when you don't.

NOTE: The Calculator provides a planning estimate. It does not replace a detailed capacity or RF design when application requirements, device behavior, spectrum availability, or site conditions require more detailed analysis.

1

Determine the Application Profile

The application profile is an inventory of the devices and applications that will use the network.

For each device category, capture:

• Application: What does the device or application do? Examples include Automated Guided Vehicle (AGV) navigation, video surveillance, handheld scanning, voice, and Supervisory Control and Data Acquisition (SCADA).
• Peak concurrent devices: How many devices of that type are expected to be simultaneously active during peak operating conditions?

  NOTE: Size the network for the expected peak operating workload, not simply the total number of deployed devices or the average traffic level. Identify how many devices are expected to be active simultaneously and what traffic each application generates during the peak operating condition.
• Downlink requirement: How much data does the application receive?
• Uplink requirement: How much data does the application transmit?
• Latency requirement: How sensitive is the application to network delay?
• Mobility: Is the device stationary, pedestrian, or vehicle-mounted?

Why Application Profiling Comes First?

Not all devices place the same demands on the network.

A handheld barcode scanner may use only a small amount of bandwidth, while a 4K surveillance camera can generate a continuous high-volume uplink stream. Similarly, an AGV performing real-time navigation may have more demanding latency requirements than a sensor that periodically reports a measurement.

Understanding the complete application profile provides the baseline for calculating network capacity and determining the required AP density.

Application Traffic Classes

NOTE: The following ranges provide general planning guidance for classifying application traffic. They are illustrative ranges, not guaranteed application requirements.

Class	Examples	Typical Range
Light	Voice, messaging, telemetry	100–300 kbps
Moderate	IoT, monitoring systems, standard video applications	500 kbps – 3 Mbps
Heavy	AR/VR, machine vision, real-time video analytics	10 Mbps or greater

NOTE: Actual bandwidth depends on application design, concurrency, video resolution, frame rate, compression, operational workload, and device behavior. Use application-specific requirements whenever they are available.

ROM Calculator Bandwidth Tiers

The Calculator uses broader planning buckets than the application traffic ranges above. These DL and UL values are what the calculator maps to your Downlink/User and Uplink/User selections.

Class	Examples	DL	UL
Light / Low	Voice, messaging, telemetry	2 Mbps	0.5 Mbps
Moderate	IoT, monitoring, standard video	20 Mbps	5 Mbps
Heavy / High	AR/VR, machine vision, real-time analytics	75 Mbps	25 Mbps

NOTE: The Calculator applies a 0.65 concurrency factor to the device count used in its capacity calculation. Do not independently apply an additional concurrency reduction when interpreting the Calculator's estimate. For deployments where the expected traffic profile does not map well to the Calculator's generic tiers, validate the result using the application-specific calculation described below.

2

Calculate the Total Throughput Demand

Calculate the aggregate throughput requirement for the defined peak operating condition.

For each device category:

Total DL demand = active devices × DL bandwidth per device

Total UL demand = active devices × UL bandwidth per device

Sum the results across all device categories.

Calculate downlink and uplink independently because enterprise applications rarely generate equal traffic in both directions.

The result is the peak throughput requirement that the network must support during the defined peak operating condition.

NOTE: Capacity planning should use the expected peak concurrent workload rather than average utilization. Designing only for average traffic can result in insufficient capacity during normal operational peaks.

3

Select the Technology and Architecture

With the peak throughput requirement established, compare it against the capabilities of the Onyx LTE and 5G platforms.

Technology selection should consider: Peak downlink throughput, peak uplink throughput, application latency requirements, device capabilities, spectrum availability, MIMO capability, expected device concurrency, mobility requirements, capacity-driven cell count, deployment architecture, and future operational requirements.

Cell capacity is determined from the LTE or 5G SA TDD tables below (frame format × bandwidth × MIMO):

APs_cap = max(⌈TotalDL ÷ CellDL⌉, ⌈TotalUL ÷ CellUL⌉)

LTE or 5G?

In many deployments, technology selection needs to be made early because GXC Onyx LTE and 5G use distinct hardware platforms.

LTE vs 5G At a Glance

	LTE	5G
Latency Characteristics	Suitable for moderate-latency enterprise applications (>50 ms)	Supports lower-latency and real-time workloads (<30 ms)
Typical Use Cases	General enterprise connectivity, telemetry, IoT	Real-time analytics, robotics, machine vision, broadcast video
Estimated Peak AP Throughput	~40 Mbps @ 20 MHz (general baseline); ~105 Mbps/cell at 20 MHz for the default Config1 configuration (up to 96 RRC users), see the table below for the full range by configuration	Up to 300+ Mbps with 4×4 MIMO @ 40 MHz, depending on deployment conditions
Channel BW options in calculator	Fixed (2×20 MHz CA reference)	20 / 40 / 100 MHz, selectable

NOTE: Actual throughput and latency depend on spectrum allocation, RF conditions, antenna configuration, device capability, mobility, interference, and overall deployment design. Use the measured throughput tables in this guide as planning references rather than guaranteed production performance.

LTE and 5G Hardware Considerations

GXC Onyx LTE and 5G use distinct hardware platforms. Selecting the appropriate technology is therefore an important early design decision.

If an existing LTE deployment later requires 5G capability, a supported 5G architecture may be available for specific configurations. This is not a standard LTE-to-5G migration model.

LTE Capacity

This section provides guidance for understanding LTE capacity on the Onyx platform.

LTE Capacity Overview

LTE capacity planning determines how much traffic the network can support and how many cells are required to meet the expected peak demand.

LTE Time Division Duplex (TDD) shares the available radio spectrum between downlink and uplink transmission. Capacity therefore depends not only on channel bandwidth and MIMO configuration, but also on how radio resources are divided between the two directions.

For LTE deployments, capacity planning should consider:

• Peak downlink throughput
• Peak uplink throughput
• Traffic direction and application behavior
• TDD Subframe Assignment (SA)
• Channel bandwidth
• MIMO configuration
• Modulation
• Carrier Aggregation (CA), where supported
• Device capability
• RF conditions
• Expected peak device concurrency

The goal is to select an LTE configuration that provides sufficient capacity in both directions while avoiding unnecessary cell deployment.

Select the LTE TDD Subframe Assignment

LTE Time Division Duplex (TDD) uses the same frequency channel for downlink and uplink traffic. The network alternates between the two directions, allocating a defined portion of radio resources to each.

The Subframe Assignment (SA) determines this DL/UL allocation.

Think of the radio channel as a road shared by traffic moving in two directions. Allocating more time to one direction increases capacity in that direction, but leaves less capacity available for the other.

Why Does it Matter?

Choosing the appropriate SA profile is an important part of LTE capacity planning.

For example, a surveillance deployment may continuously transmit video from cameras to a recording system. Selecting a heavily downlink-oriented SA profile for such a deployment would allocate insufficient airtime to uplink traffic.

Available SA Profiles

SA	DL	UL	Traffic Profile
SA1	70%	30%	Balanced traffic
SA2	85%	15%	Downlink-heavy traffic
SA6	60%	40%	Uplink-heavy traffic

Choosing an SA Profile Rule of Thumb

• Balanced traffic · General enterprise connectivity, tablets, laptops, and mixed IoT → Consider SA1 as the starting point.
• Primarily downlink traffic · Streaming video, software updates, command delivery, and content distribution → Consider SA2.
• Primarily uplink traffic · Cameras, sensors, AGVs, telemetry, and industrial monitoring → Consider SA6.

For example, a surveillance deployment may continuously transmit video from cameras to a recording or monitoring system. A heavily downlink-oriented SA would allocate insufficient airtime to this uplink traffic.

These recommendations are starting points: the final selection should be based on the calculated application traffic profile.

Application	Recommended SA
CPE routers serving broadband traffic	SA2: DL-heavy
Video streaming to displays or signage	SA2: DL-heavy
Handheld scanners downloading work orders	SA2: primarily DL
Mixed enterprise (tablets, laptops, light IoT	SA1) balanced
Push-to-talk voice with data	SA1: balanced
Video surveillance cameras	SA6: UL-heavy
Dense sensor/telemetry networks	SA6: UL-heavy
AGVs or robots reporting position and status	SA6: UL-heavy
Unknown or mixed traffic profile	SA1: starting point

NOTE: Every GXC AP operating on the same frequency channel must use the same SA configuration. SA is therefore a network-level design decision rather than an individual device setting.

Understand LTE Throughput

The following throughput tables provide measured peak throughput values observed during controlled GXC testing.

Use these values as planning references, not guaranteed production throughput.

Actual performance depends on RF signal quality, spectrum availability, interference, channel bandwidth, MIMO configuration, modulation, device capability, AP density, user concurrency, mobility, and environmental conditions.

Single/Dual Carrier Peak Expected Throughput (Mbps)

MIMO Config	Dir/Mod	SA1 10MHz	SA1 20MHz	SA2 10MHz	SA2 20MHz	SA6 10MHz	SA6 20MHz
DL 2T	DL-256QAM	48	96	66	131	39	78
DL 1T	DL-256QAM	24	48	33	66	20	39
UL 1T	UL-64QAM	14	29	7	14	18	36

Carrier Aggregation Peak Expected Throughput (Mbps)

MIMO Config	Dir/Mod	SA1 2×20MHz	SA1 2×10MHz	SA1 20+10MHz	SA2 2×20MHz	SA2 2×10MHz	SA2 20+10MHz	SA6 2×20MHz	SA6 2×10MHz	SA6 20+10MHz
DL 2T	DL-256QAM	192	96	144	262	131	197	157	78	118
DL 1T	DL-256QAM	96	48	72	131	66	98	78	39	59
UL 1T	UL-64QAM	58	29	43	29	14	22	72	36	54

IMPORTANT: These values represent measured performance under controlled conditions. They should not be interpreted as guaranteed per-user throughput or as a substitute for detailed RF and capacity engineering.

Understand MIMO, Modulation, and Channel Width

MIMO Configuration

MIMO refers to the number of spatial streams used simultaneously for data transmission or reception. Higher-order MIMO can increase throughput when supported by the AP, client device, antenna configuration, and RF conditions.

Configuration	Description	Typical Devices
DL 2T (2×2)	Two simultaneous downlink streams from the AP	Supported by many enterprise devices and CPEs
DL 1T (1×1)	One downlink stream from the AP	Basic devices or deployments where higher-order MIMO is not available
UL 1T (1×1)	One uplink stream to the AP	Common LTE UL configuration

Modulation

Modulation determines how much data can be transmitted in each radio signal. Higher-order modulations can carry more bits per transmission, but requires better RF conditions.

Channel Width

The Single/Dual Carrier table's column headers (10 MHz, 20 MHz) represent the spectrum allocated to the deployment. Wider channels generally provide greater throughput capacity when the additional spectrum is available and supported by the deployment.

Consider Carrier Aggregation

CA combines multiple LTE carriers to provide additional usable spectrum capacity to supported devices. For example, two 20 MHz carriers can provide an aggregate 40 MHz spectrum allocation.

Combination	Channel 1	Channel 2	Effective BW	Use Case
2×20 MHz	20	20	40 equivalent	Maximum capacity: use when two 20 MHz grants are available
2×10 MHz	10	10	20 equivalent	Two narrower grants bonded: same total as a single 20 MHz channel
20+10 MHz	20	10	30 equivalent	Asymmetric CA: useful when one wider and one narrower grant are available

NOTE: CA can increase available capacity when multiple supported carriers are available. Before including CA in the capacity design, verify spectrum availability, device support, AP and network configuration, and applicable deployment constraints.

LTE Capacity Design Considerations

• Use measured throughput values as planning references rather than theoretical maximums.
• Select the SA profile based on the dominant traffic direction.
• Use wider channel bandwidths where additional spectrum is available and additional capacity is required.
• Validate RF quality before relying on higher-order MIMO or modulation.
• Design AP density around expected peak operational demand rather than average utilization.
• Carefully evaluate uplink requirements for surveillance, telemetry, industrial automation, and sensor-heavy environments.
• Consider CA where additional spectrum is available and supported by the deployment.
• Validate device capabilities when determining realistic throughput.

The resulting LTE capacity design should establish the capacity-driven cell count before coverage and architecture are finalized.

5G Capacity

This section provides guidance for understanding 5G capacity on the Onyx platform.

5G Capacity Overview

5G capacity planning determines how much traffic the network can support and how many cells are required to meet the expected peak demand.

5G NR TDD uses the same spectrum for downlink and uplink transmission. The network therefore uses a slot pattern to determine how radio resources are allocated between the two directions.

For 5G deployments, capacity planning should consider:

• Peak downlink throughput
• Peak uplink throughput
• Traffic direction and application behavior
• TDD slot pattern
• Channel bandwidth
• MIMO configuration
• Modulation
• Device capability
• RF conditions
• Expected peak device concurrency
• Mobility and operational requirements

NOTE: The throughput values in this section are based on measured performance during controlled GXC field and laboratory testing using CBRS and mid-band 5G spectrum configurations. Actual performance varies with spectrum availability, RF conditions, interference, device capability, MIMO configuration, mobility, environmental obstructions, AP density, user concurrency, and overall deployment design.

Select the 5G TDD Slot Pattern

5G TDD uses the same spectrum for uplink and downlink transmission. A slot pattern determines how radio resources are distributed between the two directions.

• Downlink traffic moves from the network to devices, such as tablets, AR headsets, and handheld devices.
• Uplink traffic moves from devices to the network, such as cameras, sensors, robots, scanners, and telemetry systems.

A deployment with many cameras, sensors, or robots may require greater uplink allocation. A deployment dominated by video delivery, software downloads, or AR content may require greater downlink allocation.

5-Slot Patterns

Provides simplified scheduling behavior and predictable traffic allocation.

Pattern	DL/UL Distribution	Recommended Deployment Environment
DDDSU	74% DL / 23% UL	Downlink-heavy applications including content delivery and media distribution
DDSUU	54% DL / 43% UL	Balanced enterprise traffic and mixed-use environments
DSUUU	34% DL / 63% UL	Uplink-heavy environments including surveillance and telemetry workloads

10-Slot Patterns

Provides finer control over downlink and uplink resource allocation, for mixed enterprise workloads.

Pattern	DL/UL Distribution	Recommended Deployment Environment
DDDDDDDSUU	77% DL / 21% UL	Downlink-heavy applications such as video delivery, AR/VR, and content distribution
DSUUUUDSUU	25% DL / 72% UL	Mixed enterprise traffic including handheld devices, laptops, and general connectivity
DDSUUUUUUU	20% DL / 77% UL	Uplink-heavy environments such as surveillance, telemetry, and industrial sensors

NOTE: The DL and UL percentages represent the approximate allocation of data-carrying resources. They do not necessarily sum to 100% because TDD frame structures also include special and other non-DL/UL resources.

Choosing Between 5-Slot and 10-Slot Patterns

Consider dominant traffic direction, aggregate throughput requirement, scheduling behavior, latency requirements, synchronization and coexistence requirements, and operational simplicity.

Choose a Slot Pattern by Traffic Profile

Heavy Downlink Applications

Examples include video delivery, AR/VR, content distribution, and software distribution. Both DDDSU and DDDDDDDSUU allocate the majority of airtime to downlink.

Pattern	Split	Best when
DDDSU (5-slot)	74% DL / 23% UL	Simpler scheduling preferred, predictable DL delivery, no coexistence constraint, lower operational complexity
DDDDDDDSUU (10-slot)	77% DL / 21% UL	Maximum DL throughput ceiling required, fine-grained DL/UL balancing needed, adjacent networks require 10-slot synchronization

NOTE: DDDSU is the recommended starting point for most Heavy DL deployments. DDDDDDDSUU delivers marginally higher DL capacity and is the right choice when squeezing maximum downlink throughput from available spectrum is the priority.

Mixed-Use Applications

Examples include general enterprise connectivity, handheld devices, laptops, balanced industrial workloads.

Pattern	Split	Best When
DDSUU (5-slot)	54% DL / 43% UL	Balanced enterprise environment, operational simplicity preferred, predictable traffic allocation
DSUUUUDSUU (10-slot)	25% DL / 72% UL	Greater flexibility needed to tune DL/UL balance dynamically, adjacent networks require 10-slot synchronization

NOTE: DDSUU is the recommended starting point for most Mixed Use deployments. DSUUUUDSUU provides more fine-grained control over the DL/UL balance and suits environments where traffic patterns vary significantly across shifts or operational periods.

Heavy Uplink Applications

Examples include surveillance cameras, telemetry sensors, AGVs, and industrial monitoring. Both DSUUU and DDSUUUUUUU allocate the majority of airtime to uplink.

Pattern	Split	Best when
DSUUU (5-slot)	34% DL / 63% UL	Many devices uploading frequently, low-latency UL response needed, high device concurrency, operational simplicity preferred
DDSUUUUUUU (10-slot)	20% DL / 77% UL	Maximum raw UL throughput ceiling required, sustained bulk continuous uploads, adjacent networks require 10-slot synchronization

NOTE: For most Heavy UL deployments, DSUUU is the recommended starting point: more frequent uplink scheduling opportunities reduce per-device queuing delay, which matters with high concurrent device activity. DDSUUUUUUU delivers a higher aggregate UL throughput ceiling and is the right choice when bulk continuous upload capacity is the priority and latency tolerance is higher.

Understand 5G Throughput

The following throughput tables provide measured peak throughput values observed during controlled GXC testing.

IMPORTANT: Use these values as deployment planning references, not guaranteed per-user throughput in a shared production environment. Actual performance varies with RF quality, channel bandwidth, MIMO configuration, modulation, device capability, AP density, user concurrency, mobility, and environmental conditions.

Peak Expected Throughput (Mbps) · 10-Slot Patterns

10-slot · DDDDDDDSUU (77% DL / 21% UL): Heavy Downlink

MIMO Config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	124	263	546	685	1370
DL 2T	DL-256QAM	63	135	280	351	703
DL 1T	DL-256QAM	32	69	144	180	361
UL 2T	UL-256QAM	15	31	64	81	161
UL 2T	UL-64QAM	12	24	51	64	125
UL 1T	UL-256QAM	7	16	32	40	80
UL 1T	UL-64QAM	6	12	25	32	62

10-slot · DSUUUUDSUU (25% DL / 72% UL): Mixed Use

MIMO Config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	55	117	243	305	609
DL 2T	DL-256QAM	28	60	124	156	312
DL 1T	DL-256QAM	14	31	64	80	160
UL 2T	UL-256QAM	43	91	189	237	471
UL 2T	UL-64QAM	34	72	149	187	365
UL 1T	UL-256QAM	21	46	95	119	236
UL 1T	UL-64QAM	17	36	74	93	183

10-slot · DDSUUUUUUU (20% DL / 77% UL): Heavy Uplink

MIMO Config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	44	92	192	241	482
DL 2T	DL-256QAM	22	47	98	124	247
DL 1T	DL-256QAM	11	24	51	63	127
UL 2T	UL-256QAM	49	103	215	270	536
UL 2T	UL-64QAM	38	81	169	212	415
UL 1T	UL-256QAM	24	52	107	135	268
UL 1T	UL-64QAM	19	41	85	106	208

Peak Expected Throughput (Mbps) · 5-Slot Patterns

5-slot · DDDSU (74% DL / 23% UL): Heavy Downlink

MIMO Config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	119	253	526	660	1319
DL 2T	DL-256QAM	61	130	270	338	677
DL 1T	DL-256QAM	31	67	138	173	347
UL 2T	UL-256QAM	16	33	69	86	171
UL 2T	UL-64QAM	12	26	54	68	150
UL 1T	UL-256QAM	8	17	34	43	86
UL 1T	UL-64QAM	6	13	27	34	66

5-slot · DDSUU (54% DL / 43% UL): Mixed Use

MIMO config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	87	185	384	482	964
DL 2T	DL-256QAM	45	95	197	247	495
DL 1T	DL-256QAM	23	49	101	127	254
UL 2T	UL-256QAM	29	62	129	162	321
UL 2T	UL-64QAM	23	49	101	127	249
UL 1T	UL-256QAM	15	31	64	81	161
UL 1T	UL-64QAM	12	24	51	64	125

5-slot · DSUUU (34% DL / 63% UL): Heavy Uplink

MIMO Config	Dir/Mod	10MHz	20MHz	40MHz	50MHz	100MHz
DL 4T	DL-256QAM	55	117	243	305	609
DL 2T	DL-256QAM	28	60	124	156	312
DL 1T	DL-256QAM	14	31	64	80	160
UL 2T	UL-256QAM	43	91	189	237	471
UL 2T	UL-64QAM	34	72	149	187	365
UL 1T	UL-256QAM	21	46	95	119	236
UL 1T	UL-64QAM	17	36	74	93	183

Understand 5G MIMO, Modulation, and Channel Bandwidth

MIMO Configuration

Higher-order MIMO can increase throughput and spectral efficiency when supported by the AP, client device, antenna configuration, and RF conditions.

Configuration	Description	Typical devices
DL 4T (4×4)	Four simultaneous downlink streams from the AP	High-end devices, industrial CPEs with four receive antennas
DL 2T (2×2)	Two simultaneous downlink streams from the AP	Mid-range smartphones, most enterprise CPEs
DL 1T (1×1)	One downlink stream from the AP	Basic devices, coverage-edge scenarios