Your video call is perfectly smooth. Then a camera in another room starts uploading, and suddenly you sound like a malfunctioning robot. Your Wi-Fi icon still shows full bars. The router has not moved. What just happened?
One possible culprit is the Wi-Fi hidden node problem: two devices can communicate with the same access point but cannot reliably detect each other’s transmissions. Both may decide the channel is available and talk over each other at the receiver. Good signal strength does not prevent bad timing.
This is a small concept with a surprisingly useful lesson: a wireless network depends on who can hear whom, not just who can reach the router.
What is a hidden node in Wi-Fi?
Imagine two people in separate rooms calling out to a receptionist in a hallway. Each hears the receptionist clearly. A thick wall prevents them from hearing each other. Both wait until their own room sounds quiet, then speak. At the reception desk, their voices overlap.
Replace the people with wireless devices and the receptionist with an access point. That is the classic hidden node, or hidden terminal, scenario. “Hidden” describes a radio relationship; it does not mean the device is concealed from the network administrator.
Consider this illustrative arrangement, with both clients using the same Wi-Fi channel:
| Radio relationship | What happens |
|---|---|
| Phone and access point | They can exchange frames. |
| Camera and access point | They can exchange frames. |
| Phone and camera | They cannot reliably detect each other’s transmissions. |
| Both clients transmit together | Their frames may interfere at the access point. |
Distance, walls, antenna patterns, and differences in transmit power can all contribute. There is no requirement for a device to have a weak connection to the access point.
Why listening before talking is not enough
Wi-Fi uses carrier sensing and contention mechanisms to share airtime. During ordinary contention-based access, a station checks whether the medium is busy and follows waiting and random backoff rules before transmitting. Our CSMA/CA explainer introduces that framework.
The catch is wonderfully simple: a station can only sense what reaches its own radio. A quiet channel at the phone does not guarantee a quiet channel at the access point.
If overlapping transmissions prevent successful reception, the sender may miss the expected acknowledgment and retry. More attempts consume more airtime, leaving less time for useful traffic. For ordinary unicast exchanges this involves acknowledgments; aggregated traffic can use block acknowledgments instead of a separate ACK for every frame.
A missing acknowledgment does not identify the cause. Interference, weak reception, and other problems can also produce retries. Hidden nodes are one explanation, not a diagnosis printed inside every failed packet.
Why full signal bars can fool you
The Wi-Fi icon generally summarizes received signal strength using a device-specific scale. It is not an airtime meter, a measurement of the access point’s reception of your uploads, or a promise of internet speed.
Even a useful radio measurement needs context. Signal-to-noise ratio compares the desired signal with background noise; channel utilization describes how busy the channel appears at the measuring radio. Neither reveals every relationship between every pair of devices.
That is why “move closer until the bars improve” can miss the problem. In our example, both clients already hear the access point well. Their missing relationship is with each other.
A practical test: change the conversation
Start by separating wireless performance from internet performance. A local transfer to a computer connected by Ethernet avoids involving an external speed-test server or your broadband connection.
Here is a useful troubleshooting sequence, rather than a definitive hidden-node detector:
- Establish a baseline. Confirm the test clients use the intended access point, band, and channel. Test each client’s upload separately to the wired computer, keeping the setup consistent.
- Introduce overlap. Run both uploads together. Record aggregate throughput, responsiveness, and available retry statistics. Some slowdown is normal because the clients share airtime.
- Change the radio relationship. Move one client so the two are less obstructed while keeping their access-point connections reasonably comparable. Repeat the individual and simultaneous tests.
- Look for a repeatable pattern. A disproportionate performance collapse under overlap, followed by improvement after repositioning, supports further investigation. It does not rule out interference, rate changes, or other variables affected by the move.
For a deeper investigation, use access-point telemetry and an over-the-air capture. A normal Ethernet capture cannot show all the wireless control frames and retries. Capture location matters too: the monitoring radio may itself be unable to hear one of the clients, and undecodable collisions may never appear as readable packets.
RTS/CTS: ask the receptionist to reserve the room
Wi-Fi has a mechanism that can help: Request to Send / Clear to Send, or RTS/CTS.
In the classic protected exchange, a sender first transmits a short RTS. The receiver answers with CTS. Nearby stations that receive the relevant reservation information defer for the indicated period using virtual carrier sensing, commonly described through the Network Allocation Vector, or NAV.
The clever part is the receiver’s voice. A camera that cannot hear the phone’s RTS may still hear the access point’s CTS and hold back. The reservation can protect the subsequent data-and-acknowledgment exchange.
Protection costs airtime, however. Adding control exchanges to traffic that was already working well can reduce throughput. RTS frames can also collide, and a station that cannot receive the reservation cannot act on it.
Some equipment exposes an RTS threshold, typically related to frame size. Defaults, adaptive behavior, and which transmissions a setting affects depend on the product. Changing an access point’s threshold does not automatically make every client protect its uploads. Check the vendor documentation, change one parameter at a time, and compare before and after.
Fix the layout before hunting for a magic setting
Often, the most useful intervention is a better radio layout. Reposition clients or the access point and test again. Our guide to Wi-Fi obstacles and placement provides useful context for thinking about walls and furniture.
Where practical, wire a busy stationary device such as a camera. Alternatively, distribute clients across suitably planned access-point radios and channels. Merely creating another network name on the same radio does not create another pool of airtime.
Be deliberate about extra hardware. A Wi-Fi range extender can improve reach, but its wireless backhaul also needs airtime. Adding access points without a channel plan can create more competition. “More Wi-Fi” is not automatically better Wi-Fi.
Doesn’t modern Wi-Fi solve this?
Modern Wi-Fi can coordinate simultaneous transmissions. OFDMA, introduced with Wi-Fi 6, lets an access point assign different frequency resources to compatible clients within a transmission opportunity. That is very different from two unscheduled transmissions colliding at a receiver.
Nevertheless, scheduled multiuser operation does not make every transmission scheduled or remove the need to obtain channel access. Support, configuration, traffic, and neighboring networks still matter. A newer Wi-Fi generation is useful technology, not immunity to an awkward building layout.
The next time someone says, “But I have full bars,” ask a better question: can the devices sharing this channel hear each other well enough to take turns? Sometimes the network’s problem is two perfectly good connections having one badly coordinated conversation.
References and further reading
- Cisco: Hidden Nodes Interference — a concrete hidden-node example in mesh backhaul.
- Cisco Meraki: Analyzing Wireless Packet Captures — control frames, acknowledgments, and optional RTS/CTS protection.
- Cisco: Validate Wi-Fi Throughput Testing and Monitoring Guide — local testing and RF troubleshooting.
- HPE Aruba Networking: WLAN SSID Profile — an example of product-specific RTS threshold behavior.
- Cisco Meraki: Signal-to-Noise Ratio and Wireless Signal Strength — signal quality and retransmission costs.
- Cisco Meraki: Access Point Range and Signal Strength Maximization — placement, obstacles, and channel choices.
- Cisco: Wi-Fi 6 OFDMA Resource Unit Allocations and Mappings — coordinated multiuser operation and channel contention.