Introduction
While working for a company that specialized in alarm systems and deployed telephony solutions—primarily Voice over Wi-Fi (VoWiFi)—I was responsible for designing and maintaining the wired and wireless networks that supported these critical services. The wireless infrastructure in place was based on legacy Meru/Fortinet controllers and access points (APs). For those familiar with Meru Networks, their key differentiator was the use of Single Channel Architecture (SCA), where all APs broadcast on the same channel and share a virtual BSSID MAC address. This design causes clients to see the network as a single AP, with roaming decisions managed centrally by the controller. This essay outlines the limitations of SCA in modern networks and how I successfully transitioned a customer to a Multi Channel Architecture (MCA) to resolve persistent performance issues.
Problem
A long-standing customer contacted us once again to report poor Wi-Fi performance, including one-way audio and other voice-related issues. Their network was running on a FortiWLC controller (a rebranded Meru controller) with Fortinet-branded Meru APs deployed in an SCA configuration. Although the customer had experienced intermittent issues in the past, the frequency and severity of complaints had increased, prompting the need for a more permanent solution.
Upon arriving onsite, I was able to witness the reported issues. Packet captures revealed a high number of retries and elevated channel utilization. The network had been in place for over five years, and due to budget constraints, a full infrastructure upgrade was not feasible at the time. I needed to find a workaround that would improve performance and extend the network’s viability for at least another two to three years.
Solution
To begin, I conducted a thorough diagnostic assessment. I checked for APs with high client counts, but the maximum number of clients per AP was around 15–20—well within acceptable limits for environments without high-bandwidth applications like video streaming. However, further analysis of the packet captures revealed transmissions from clients located in distant care units, far from my capture location. This indicated that clients across the campus were sharing the same channel.
My CWNP training had taught me the fundamentals of Wi-Fi operation, including the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) mechanism. In an SCA deployment, all APs operate on the same channel, meaning the available airtime is shared among all clients and APs within range. This results in significant co-channel interference (CCI), especially when over 100
clients are contending for the same medium. The consequences include increased latency, jitter, and retransmissions—conditions destructive to voice and real-time applications.
Although the AP placement was not ideal, it was serviceable. Given the limited budget, relocating or adding APs was not an option. Fortunately, the FortiWLC platform supported a native cell mode, allowing me to reconfigure the network for MCA. I began by piloting the new configuration in a specific wing of the campus. I developed a manual channel plan to control the RF environment and set transmit power ranges to 8–11 dBm on 2.4 GHz and 13–16 dBm on 5 GHz. These settings provided a balance between coverage and interference mitigation.
Before deploying the changes, I consulted with my colleagues responsible for configuring the VoIP phones. I’ve heard before that the smartphones had been optimized for SCA, with limited channel scanning and adjusted roaming thresholds. We reconfigured the phones to support MCA, enabling them to scan all channels and behave on “standard” roaming thresholds.
Once the client devices were updated, I implemented the MCA configuration and conducted a post-deployment validation using Ekahau. The survey confirmed proper channel allocation, acceptable transmit power levels, and a clean RF spectrum. The Ekahau Sidekick allowed me to perform simultaneous spectrum analysis, ensuring no hidden interference sources were present. The deployment was successful, and after a few weeks of monitoring, the customer reported a complete resolution of the performance issues. This success led to the rollout of the MCA configuration across the entire campus.:
Conclusion
My experience with SCA has shown that while it may have been beneficial in earlier Wi-Fi deployments, it no longer meets the demands of modern networks. Advances in AP and client capabilities have made MCA the preferred architecture, offering better scalability, performance, and reliability. Migrating from SCA to MCA requires careful planning, especially when dealing with legacy client configurations and budget constraints. However, with a methodical approach and a solid understanding of wireless fundamentals, it is possible to deliver a significantly improved user experience without a complete infrastructure overhaul.