Who This Is For: Network engineers evaluating WiFi 6 for 6GHz industrial wireless bridges, technical decision-makers considering an 802.11ac-to-802.11ax upgrade, and anyone who wants the details on OFDMA and TWT in real industrial PtMP deployments.
Core Issue: 802.11ac (WiFi 5) hits a wall in high-density PtMP environments — CSMA/CA contention collapse, limited channel width, no uplink MU-MIMO, and no effective power saving for sensor networks. 802.11ax (WiFi 6) was built to solve exactly these problems.
Key Conclusions: WiFi 6 combined with the 6GHz band delivers 1500+ Mbps aggregate throughput in 6GHz industrial wireless bridge deployments — roughly 3× what 802.11ac products deliver (500 Mbps). OFDMA is the game-changer for high-density PtMP scenarios, and when paired with the iPoll 3 proprietary protocol, it enables deterministic low-latency scheduling even with 20+ CPEs. The jump from 802.11ac to 802.11ax is the single biggest performance leap the industrial wireless bridge market has seen.
To understand what WiFi 6 actually brings, you need to see where the bottlenecks were.
802.11n (WiFi 4) brought MIMO and 40MHz channel bonding, pushing single-link rates from 54 Mbps to 300 Mbps. The idea was simple — add more antennas, use more spectrum.
802.11ac (WiFi 5) went further on 5GHz: 80MHz and 160MHz channel widths, 4×4 MIMO, 256-QAM, pushing PHY rates to 1.73 Gbps (4×4, 160MHz). But the MAC layer was still running CSMA/CA contention — every device competes for the same channel. More devices means more collisions, more backoff, and effective throughput that tanks fast. In a 20+ CPE industrial PtMP deployment, CSMA/CA effective throughput can drop to 20-30% of the PHY rate.
802.11ax (WiFi 6) changes the MAC layer fundamentally: OFDMA (Orthogonal Frequency Division Multiple Access) splits a channel into multiple Resource Units (RUs), and the AP assigns different RUs to different devices simultaneously. This is a paradigm shift from time-domain contention to frequency-domain scheduling.
Here’s how three generations compare side by side:
| Feature | 802.11n (WiFi 4) | 802.11ac (WiFi 5) | 802.11ax (WiFi 6) |
|---|---|---|---|
| Release Year | 2009 | 2013 | 2019 |
| Frequency Bands | 2.4/5GHz | 5GHz | 2.4/5/6GHz (WiFi 6E) |
| Channel Width | 20/40MHz | 20/40/80/160MHz | 20/40/80/160MHz |
| MIMO | 1-4×4 | 1-4×4 (DL MU-MIMO) | 1-8×8 (DL+UL MU-MIMO) |
| Modulation | 64-QAM | 256-QAM | 1024-QAM |
| Subcarrier Spacing | 312.5 kHz | 312.5 kHz | 78.125 kHz |
| OFDMA | No | No | Yes (DL+UL) |
| TWT | No | No | Yes |
| BSS Coloring | No | No | Yes |
| Max PHY Rate (2×2) | 300 Mbps (40MHz) | 867 Mbps (80MHz) | 2.4 Gbps (160MHz) |
| Typical TCP Throughput (2×2) | ~150 Mbps | ~500 Mbps | ~1500 Mbps |
One technical detail worth noting: 802.11ax shrinks subcarrier spacing from 312.5 kHz down to 78.125 kHz — a 4× reduction. Same bandwidth, 4× more subcarriers. Each subcarrier carries less data, but the OFDM symbol duration goes from 3.2μs to 12.8μs. Longer symbols mean stronger resistance to multipath fading — a real advantage in industrial NLOS and reflective environments. The narrower subcarriers also let you slice frequency resources more finely across multiple users.
The 6GHz band is the ideal physical-layer home for 802.11ax: clean spectrum (noise floor -100 dBm to -110 dBm), plenty of channels (30+ non-overlapping 20MHz channels), and no DFS restrictions. When WiFi 6 technical advantages stack on top of 6GHz spectrum advantages, industrial wireless bridge performance jumps from “good enough” to “abundant.”
Best way to understand OFDMA is by analogy:
OFDM (legacy, 802.11a/n/ac) works like a single-lane road: at any given moment, only one device can use the full channel. Even if a CPE only needs a tiny slice of spectrum, every other CPE has to wait their turn. Imagine a 4-lane highway where only one car is allowed to drive at a time — everyone else queues at the toll booth.
OFDMA (802.11ax) works like a multi-lane highway: the AP divides the channel into multiple Resource Units (RUs), each containing a specific set of subcarriers. In a single TXOP, the AP can assign different RUs to different CPEs and they all transmit simultaneously. Think of it as splitting that highway into dedicated lanes so different vehicles can travel at the same time.
802.11ax defines 6 RU sizes with different subcarrier counts and data capacities:
| RU Size | Subcarriers | Max RUs in 20MHz | Max RUs in 80MHz | Typical Use Case |
|---|---|---|---|---|
| 26-Tone RU | 26 | 9 | 37 | Low-bandwidth IoT/sensors (~0.5-1 Mbps) |
| 52-Tone RU | 52 | 4 | 18 | Low-rate data acquisition |
| 106-Tone RU | 106 | 2 | 8 | Medium-rate surveillance/video |
| 242-Tone RU | 242 | 1 | 4 | Full-channel high-speed (~20MHz equivalent) |
| 484-Tone RU | 484 | — | 2 | 40MHz channel bonding |
| 996-Tone RU | 996 | — | 1 | Full 80MHz channel |
Why this matters for industrial deployments: In a typical PtMP campus network, CPE bandwidth needs are uneven — a surveillance camera CPE might need 20-50 Mbps uplink, a sensor CPE needs 0.1-1 Mbps, and a voice CPE needs low latency but low bandwidth. With OFDMA, the AP can allocate a 242-tone RU (high bandwidth) to the video CPE and a 26-tone RU (low bandwidth) to the sensor CPE in the same transmission slot — no sequential contention required. That kind of differentiated frequency-domain scheduling is impossible with OFDM.
Take a typical industrial park PtMP deployment — 1 base station + 15 CPEs with mixed traffic (6 video feeds + 5 data links + 4 sensors):
| Performance Metric | Legacy OFDM (802.11ac) | OFDMA (802.11ax) | Improvement |
|---|---|---|---|
| Channel Access | CSMA/CA contention | AP scheduled | Deterministic |
| Simultaneous Users | Not supported | Up to 37 parallel | 37× |
| Small-Packet Efficiency | Very low (high contention overhead) | High (RU allocated on demand) | 2-5× |
| Channel Utilization | ~40-60% | ~70-90% | ↑ 50% |
| Mixed-Traffic Aggregate Throughput | ~300-400 Mbps | ~600-900 Mbps | ↑ 100-125% |
| Worst-Case Latency | 50-200 ms (collision backoff) | 10-30 ms | ↓ 80-85% |
Key insight: OFDMA gains are most dramatic with small packets. Industrial networks are full of small-packet traffic — sensor readings (tens to hundreds of bytes), Modbus/Profibus control protocol messages (tens to hundreds of bytes), VoIP packets (~200 bytes). In legacy OFDM, each small packet still occupies the full channel for transmission, and the preamble + contention overhead dwarfs the actual payload. OFDMA bundles multiple small packets into a single transmission opportunity using frequency-domain multiplexing, slashing the per-bit overhead.
MU-MIMO separates users in the spatial domain. The AP’s multi-antenna array forms different beams, letting it talk to (or listen to) multiple CPEs on the same time-and-frequency resource. 802.11ax extends this to both directions:
Industrial traffic is overwhelmingly uplink — surveillance feeds going to the NVR, sensor readings heading to the control server, PLC status updates flowing upstream. Uplink MU-MIMO lets the AP schedule multiple CPEs to transmit in the same TXOP, dramatically improving uplink throughput and cutting uplink latency.
MU-MIMO and OFDMA work together, not instead of each other. In one TXOP, the AP uses OFDMA to assign different RUs to different groups, then layers MU-MIMO spatial streams on each RU. This “frequency + spatial” joint scheduling is one of 802.11ax’s biggest technical advantages.
In traditional 802.11, before transmitting, a CPE runs physical carrier sensing (CCA). If detected signal power exceeds the CCA threshold (typically -82 dBm), the CPE considers the channel busy and waits. The problem is — that signal might come from a neighboring AP that doesn’t interfere with this CPE’s transmission at all (different AP association). This “over-protection” kills spatial reuse in dense deployments.
BSS Coloring adds a 6-bit “color” field (63 possible values). Every frame the AP sends carries its color. When a CPE detects a signal:
In industrial parks or multi-AP co-located setups, BSS Coloring makes a real difference. Picture 4 YNW 6-90ax sector base stations on one rooftop covering 360°. Give each one a different color, and CPEs in one sector won’t falsely treat adjacent sector traffic as contention.
Here’s the full industrial wireless bridge comparison between 802.11ax (YNW 6ax) and 802.11ac (YNW 6ac):
| Dimension | YNW 6ac (802.11ac) | YNW 6ax (802.11ax) | Delta |
|---|---|---|---|
| Max aggregate throughput | 500 Mbps | 1500+ Mbps | 3× |
| Max PHY rate (2×2) | 867 Mbps (80MHz) | 2.4 Gbps (160MHz) | 2.8× |
| Channel width | 5/10/20/40/80MHz | 20/40/80/160MHz | +160MHz support |
| Max modulation | 256-QAM (8-bit/sym) | 1024-QAM (10-bit/sym) | +25% spectral efficiency |
| Multiple access | OFDM + CSMA/CA | OFDMA + scheduled | Far better multi-user efficiency |
| MU-MIMO | Downlink only (4 users) | Uplink + downlink (8 users) | New uplink capability |
| Subcarrier spacing | 312.5 kHz | 78.125 kHz | 4× longer symbol, stronger against multipath |
| OFDM symbol duration | 3.2 μs | 12.8 μs (13.6 μs with CP) | Better NLOS performance |
| TWT power saving | Not supported | Supported | Key for industrial IoT |
| BSS Coloring | Not supported | Supported | Better dense deployment efficiency |
| WPA3 security | Optional | Mandatory | SAE/OWE stronger encryption |
| Frequency range | 5.85-6.45GHz | 5.9-7.125GHz | Double the spectrum |
| Ethernet port | 1× Gigabit RJ45 | 2.5G PoE IN + Gigabit PoE OUT | No wired bottleneck |
| PoE | 24VDC passive PoE | 802.3at PoE (55VDC) | Standard PoE, better compatibility |
| Ingress protection | IP65 | IP67 | Better outdoor protection |
Expected TCP throughput for the YNW 6ax series in PTP mode (line of sight, 6GHz):
| Distance | Channel Width | Modulation (MCS) | Expected TCP Throughput | vs 6ac at Same Distance |
|---|---|---|---|---|
| 0-1 km | 160 MHz | 1024-QAM (MCS 11) | 1400-1500 Mbps | 500 Mbps → ↑ 200% |
| 1-2 km | 160 MHz | 1024-QAM (MCS 10) | 1200-1400 Mbps | 450-500 Mbps → ↑ 170% |
| 2-4 km | 80 MHz | 256-QAM (MCS 9) | 700-900 Mbps | 350-450 Mbps → ↑ 100% |
| 4-8 km | 40 MHz | 64-QAM (MCS 7) | 300-500 Mbps | 200-350 Mbps → ↑ 50% |
| 8-15 km | 20 MHz | QPSK (MCS 1-3) | 80-200 Mbps | 50-150 Mbps → ↑ 40% |
Factors that affect real-world throughput:
Profile: Short distance, very high throughput. Think data center interconnect, wireless backup between core switches, 4K/8K video surveillance aggregation.
Recommended kit: YNW 6-23ax (23dBi, ~10° beam) or YNW 6-20ax (20dBi, ~15° beam) on 160MHz channels. Expect 1400-1500 Mbps sustained. The 2.5G Ethernet port ensures no wired bottleneck.
vs 6ac: Same conditions, 6ac gives 450-500 Mbps. WiFi 6 delivers ~200% more. Where you needed 3× 6ac links for 1.5 Gbps, now one 6ax link does it.
Profile: 1 base station + 15-30 CPEs, mixed traffic (video + data + sensors), real-time streaming and deterministic control signaling required.
Recommended kit: Base station: YNW 6-90ax (17dBi sector, 90° × 20°). CPEs: YNW 6-20ax (within 2km) or YNW 6-23ax (2-4km). OFDMA + iPoll 3 gives double-layer scheduling — iPoll 3 decides which CPE transmits when, OFDMA decides which frequency resources that CPE uses.
vs 6ac: At 20 CPEs with mixed traffic, 6ax delivers 600-900 Mbps aggregate. 6ac in the same scenario does 300-400 Mbps.
Profile: Longer distances, still needs decent throughput. Mine-to-control-room video, oil field data aggregation, cross-campus trunking.
Recommended kit: YNW 6-23ax (23dBi antenna) on 20-40MHz channels. The 23dBi gain is the highest available in WiFi 6 6GHz gear. Combined with OFDMA and 1024-QAM spectral efficiency at narrow channel widths, expect 300-500 Mbps at 8km.
vs 6ac: 6ac (6-20ac, 20dBi) at 15km does about 150-200 Mbps on 40MHz. 6ax (6-23ax) at the same distance does 200-400 Mbps — 50-100% improvement, driven by 3 dB higher antenna gain and 1024-QAM efficiency.
WiFi 6 (802.11ax) is the standard itself — it defines OFDMA, 1024-QAM, TWT, and all the other technical features. WiFi 6E is just WiFi 6 extended into the 6GHz band — same standard, different frequency. WiFi 6E hardware needs 6GHz-capable RF front-ends (PA, LNA, filters) that regular 2.4/5GHz WiFi 6 gear doesn’t have. The YNW 6ax series is WiFi 6E.
No — they operate at different layers and work together. OFDMA lives in the 802.11ax PHY and MAC resource scheduling layer — it handles allocating RUs in the frequency domain. iPoll 3 lives in the data scheduling layer — it decides which CPE gets a transmit opportunity and when. They’re complementary: iPoll 3 decides “who” and “when,” OFDMA decides “which frequency resources.” The YNW 6ax series runs OFDMA on top of the iPoll 3 scheduling framework for joint time+frequency scheduling.
6ac and 6ax use different wireless chipset platforms (6ac: QCA 9563 + QCA 9882; 6ax: 802.11ax chipset) and different power standards (24VDC passive PoE vs 802.3at PoE). So you replace all wireless units. On the wired side, check whether your switch supports 802.3at PoE and 2.5G ports. If not, you’ll need PoE injectors or a switch upgrade. That said, if your 6ac network is running fine and you don’t need more than 500 Mbps, there’s no rush.
As of mid-2026: US (FCC) opened 5.925-7.125 GHz (1200 MHz) for unlicensed use. EU (ECC) opened 5.945-6.425 GHz (480 MHz). China has allocated parts of 5.925-7.125 GHz for industrial wireless applications. Japan, South Korea, Australia, and others have also completed regulatory openings. Specific frequency ranges and power limits vary by country — check local regulations before deploying.
The full MCS 0-11 rates for 160MHz are listed in section 3 above. For 20/40/80MHz, roughly halve the rate for each step down — 80MHz is about half of 160MHz, 40MHz is about a quarter, 20MHz is about an eighth. For the exact full table covering all bandwidth and GI combinations, see the YNW 6ax series product page specs.
About the YNW 6ax Series: The YNW 6GHz 802.11ax wireless bridge series includes the YNW 6-20ax (20dBi, 15° beam, optimal 2km), YNW 6-23ax (23dBi, 10° beam, optimal 4km), YNW 6-90ax (17dBi sector, 90° coverage, PtMP base station), and YNW 6ax (external antenna model, flexible antenna options). All models deliver 1500+ Mbps aggregate throughput, 160MHz channel support, iPoll 3 protocol, 2.5G Ethernet, and IP67 protection. Designed for high-capacity backhaul, dense PtMP campus coverage, and long-distance industrial backhaul.
References: