The 900x Gap the Cipher Cannot Close

The 900x Gap the Cipher Cannot Close

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

ADA · CALCULATION AUDIT

The 900x Gap the Cipher Cannot Close

The same encrypted smart meter reveals wildly different things depending on how often it samples. One reading per second lets individual appliance signatures — a kettle, a refrigerator compressor cycle, a television — become distinguishable, piling up roughly 2.6 million samples per meter each month. Drop to one reading every 15 minutes and about 2,880 samples still serve time-of-use billing while staying too coarse to name most appliances. The chapter’s point is that the privacy protection is the sampling rate, not the cipher — so this audit measures the gap between those two rates and asks whether encryption can ever close it.

Companion to the chapter Privacy vs Security in IoT — every number here comes from that chapter.

See the relationship before changing it

The figure reads from left to right. The blue card is sample interval. The middle card applies the page rule. The green card is monthly samples. Walk the arrows once: set the input, apply the rule, then read the result with its unit.

Walk the arrows. Encryption does not reduce how many detailed readings an authorised system can inspect.

Derive the baseline in four named moves

1. 1

   Name the input. The chapter baseline is 15 min.

2. 2

   Name the relationship. samples = 30 days x 24 h x 60 min / interval

3. 3

   Substitute with units. 30 x 24 x 60 / 15 = 2,880 samples

4. 4

   Read the result. Keep the unit beside the value. Use it only inside the technical boundary on this page.

Predict, then change sample interval

Try Predict the direction of samples = 30 days x 24 h x 60 min / interval. Test another sample interval, then compare monthly samples.

Sample interval

15 min

Chapter baseline

Monthly samples

Observe Encryption does not reduce how many detailed readings an authorised system can inspect. Reset sample interval to 15 and compare monthly samples.

Explain Encryption does not reduce how many detailed readings an authorised system can inspect.

Check yourself

What should you do before trusting a moved-control result?

Answer: Predict its direction, apply the shown relationship, keep the units, and reset to the worked baseline.

What does this small model leave out?

Answer: Only sample interval moves here. Field effects named in the technical boundary stay fixed.

Honesty note. This teaching model changes only sample interval. It is not a field prediction.

TryWith Check derivation, contrast one reading per second with one reading every 15 minutes across 30 days.

ObserveThe sample-count evidence separates roughly 2.6 million fine-grained records from about 2,880 coarse records, a near-900x gap.

ExplainEncryption hides values in transit but preserves timing and volume; reducing cadence removes observations that a cipher cannot make less frequent.

Ada: The smart-meter example rests on two sample counts — “roughly 2.6 million” per month at one sample per second, and “about 2,880” per month at one sample per fifteen minutes. Those two numbers are the whole argument for minimization, so let me confirm them and then measure the gap between them. I will use a 30-day month as the stated, illustrative basis.

• One sample per second: 1 x 60 x 60 x 24 x 30 = 2,592,000 samples — the chapter’s “roughly 2.6 million”.
• One sample per 15 minutes: (60 / 15) x 24 x 30 = 4 x 720 = 2,880 samples — the chapter’s “about 2,880”.

Both check out. Now the ratio of the two collection rates:

2,592,000 / 2,880 = 900

And that 900 is not arbitrary: fifteen minutes is 15 x 60 = 900 seconds, so the fast meter records one reading everywhere the slow meter records 1/900 of one. The fine-grained stream carries 900 times more readings per meter — and every one of those extra readings is where appliance-level inference lives.

The design point lands hard once the ratio is explicit: end-to-end encryption changes this 900x by a factor of exactly one, because the authorized utility decrypts and analyzes the data anyway. Only the sampling rate moves the number that matters. Whatever the meter never records at one-second resolution cannot later be breached, subpoenaed, or inferred — which is why the coarsest rate that still serves billing, not the strength of the cipher, is the real privacy control here.

Technical boundaries
Metadata inference, retention and secondary-use policy, missing readings, billing utility, and attack accuracy sit beyond this fixed-cadence sample count.

Work the audit first, then check the displayed derivation.

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