Companion to Tokenisation of Future 6G Assets as a Strategy to Improve ARPU for Operators. Every assumption is an input — all on a per-hour, typical urban site basis. A token sells only its asset’s spare capacity: as your own traffic (subscribers × data per sub) or your own GPU use rises, the sellable spare shrinks — token income can then rise only through the token’s base value. (v3 — occupancy-linked, minimum–central range, fully linked) Demand splits across eMBB, URLLC, mIoT slices (mix from the RANSlice twin). All figures in ₹.
Range (v3): every figure in the revenue table is computed live from the inputs — nothing is hard-coded. Each stream is computed twice, at the minimum assumptions and at the central assumptions (the paper’s Table 1 pair), and shown as a band. Change any assumption — minimum or central — and its end of every band moves.
Occupancy rule (v2): each token sells only its asset’s spare capacity. Radio spare falls as own traffic grows (subscribers × GB/sub, measured against the site type’s reference load); compute spare falls as the GPU pool’s own use rises. With occupancy accounted for, token income rises only by raising the token’s base value — never by assuming spare that own traffic has already taken. Base-price anchor: Delhi 3.3 GHz cleared at ≈₹40 crore/MHz in the 2022 auction; amortised over the licence area and the sellable window (≈1,500 h/yr of the 20-year term) that is ≈₹9/MHz·hr·km². The adopted base prices — urban ₹5, suburban ₹2.5, rural ₹1 — sit at or below that anchor and scale down with circle prices.
Slice shares re-weight the Stream-1 price (URLLC premium for latency, mIoT discount for bulk). Stream 2 is the thinner-evidence stream: GPU utilisation and VM pricing are reasonable assumptions, not measured data. Personal VMs complement handset AI (needed anyway for latency/privacy/offline) and are billed separately. All prices are assumed until a registry publishes cleared prices — the claim is the margin of clearance.