Why Channels 15, 20, 25, and 26 Land in the Gaps

Ada re-derives this chapter’s own numbers step by step, at full precision

ADA · CALCULATION AUDIT

Why Channels 15, 20, 25, and 26 Land in the Gaps

A Zigbee network on channel 20 runs fine until an access point is retuned near Wi-Fi channel 11, retries climb, and the fix is to move the PAN to channel 25 rather than turn up power. That works because an 802.15.4 channel sits at 2405 + 5 x (k - 11) MHz, so channels 15, 20, 25, and 26 each land in a ~3 MHz seam between the Wi-Fi blocks, while a 1 mW radio faces a 20-to-30 dB deficit against a 100 mW-to-1 W access point. This audit computes every centre frequency and asks whether shifting a few megahertz into a gap really beats shouting, or is only a shortlist entry until on-site packet loss confirms it.

Companion to the chapter 802.15.4 Coexistence — every number here comes from that chapter.

Ada: The chapter names channels 15, 20, 25, and 26 as the classic candidate set and puts Wi-Fi 1/6/11 at 2412, 2437, and 2462 MHz. Those are not arbitrary picks - let me compute the centre frequencies and show each candidate lands in a Wi-Fi gap.

• An 802.15.4 2.4 GHz channel k sits at 2405 + 5 x (k - 11) MHz: channel 15 is 2405 + 5 x 4 = 2425 MHz, channel 20 is 2405 + 5 x 9 = 2450 MHz, channel 25 is 2405 + 5 x 14 = 2475 MHz, and channel 26 is 2480 MHz.
• A 20 MHz Wi-Fi channel spans 20 / 5 = 4 of the 5-MHz-spaced 802.15.4 channels, so Wi-Fi 1 (2401-2423) blankets 802.15.4 channels 11-14, Wi-Fi 6 (2426-2448) covers 16-19, and Wi-Fi 11 (2451-2473) covers 21-24 - matching the chapter’s table.
• Channel 15 at 2425 MHz falls in the 2423 to 2426 gap, channel 20 at 2450 MHz in the 2448 to 2451 gap, and channels 25 and 26 sit above 2473 MHz. Every candidate lands in a ~3 MHz seam between the Wi-Fi blocks.

The power arithmetic explains why dodging beats shouting: 10 x log10(100 / 1) = 20 dB and 10 x log10(1000 / 1) = 30 dB, so a 1 mW radio next to a 100 mW-to-1 W access point runs a 20-to-30 dB deficit. When you cannot add 100-to-1000x of power, shifting a few megahertz into a spectral gap is the only lever left - but the seams are only ~3 MHz wide, so the candidate is a shortlist entry until on-site packet loss confirms it.

Every number above is taken from the chapter’s own material and re-derived step by step.