Core Networking · Study deck

Collision Mitigation Strategies

Imagine two trackers appearing with the same name.

Packet Pete is your guide for this deck.

collisionbirthday

After studying this chapter

Learning objectives

You will be able to:

Explain: If recovery becomes ordinary load, change the dominant variable: increase the value space, partition the domain, coordinate assignment, reduce offered traffic, schedule critical messages, stage reconnects, or add capacity supported by the protocol.
Explain: Independent generation can remove a central bottleneck, but it needs enough value space for the chosen risk target and a credible duplicate-detection path.
Explain: Duplicate-address repair can require discovery, reassignment, route or cache updates, application reconciliation, and audit logging.
Explain: The second draws the possible tag pairs and tests the name maker with a known match.

Major section

Start With the Failure, Not the Protocol · In 60 Seconds

Reopen the design when the group size, traffic, or attacker model changes.

IPv6 is a network addressing system with a very large address space.
A wait-and-retry rule can spread radio messages.
Unique assignment helps device names.
Listening and waiting can help radio traffic.

Major section

Identity Assignment Patterns · Choose the Probability Target Explicitly

A trusted allocator selects a value that its current state says is unused.

This removes the independent-with-replacement birthday model while the allocator state is correct.
Independent generation can remove a central bottleneck, but it needs enough value space for the chosen risk target and a credible duplicate-detection path.
The approximation gives.

Major section

Checkpoint: Evidence Before Release · Recovery Has a Load Limit

Recovery consumes the same constrained resources as useful work.

Duplicate-address repair can require discovery, reassignment, route or cache updates, application reconciliation, and audit logging.
Media recovery can require acknowledgements, backoff, retransmission, longer wake time, and application timeout handling.
If recovery becomes ordinary load, change the dominant variable: increase the value space, partition the domain, coordinate assignment, reduce offered traffic, schedule critical messages, stage reconnects, or add capacity supported by the protocol.

Major section

Checkpoint: Recovery Is Capacity · The Birthday Problem in IoT Networks

Detection does not make a collision free.

Eventual success can still miss latency, energy, or correctness budgets.
Persistent recovery is an escalation signal, not normal background noise.
The second draws the possible tag pairs and tests the name maker with a known match.
Also name the decision that a duplicate could spoil.

Deck summary

Key takeaways

Reopen the design when the group size, traffic, or attacker model changes.

A trusted allocator selects a value that its current state says is unused.
Recovery consumes the same constrained resources as useful work.
Detection does not make a collision free.

Retrieval practice

Recall check 1 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q1A fleet has duplicate device IDs, overlapping radio frames, and a second collision wave when all devices retry together; how should these failures be classified?

AThree: duplicate identity, airtime overlap, and correlated retry

BClassify the three failures as cloud ingestion bottlenecks

CTreat one successful packet as proof both are solved

DClassify the three failures as address allocation exhaustion

Show answer

Answer: A Names, airtime, and shared retry timing need separate controls.

Retrieval practice

Recall check 2 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q2Which controls separately address duplicate identity, airtime overlap, and correlated retry?

AIgnore duplicate IDs, airtime overlap and shared retries because a gateway exists.

BIncrease identifier length as if it created more radio airtime capacity.

CReassign IDs, separate airtime and randomize capped retries after shared outage.

DUse retry backoff to repair records already merged under a duplicate ID.

Show answer

Answer: C Each control targets a different collision mechanism.

Retrieval practice

Recall check 3 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q3Two redundant provisioning servers may operate during failover. Which control most directly prevents them from issuing the same local identifier?

AIncrease radio backoff after failures to prevent duplicate server allocations.

BUse atomic shared allocation or disjoint ranges across provisioning servers.

CUse a stronger antenna as the primary duplicate-identifier control.

DCount average devices alone instead of coordinating failover allocations.

Show answer

Answer: B Coordination or disjoint namespaces changes the assignment process so the servers cannot legitimately issue the same value.

Retrieval practice

Recall check 4 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q4An IPv6 node receives evidence during DAD that its tentative address is already in use. What follows from RFC 4862?

AAssign the tentative address after one ping because the probe establishes that the identifier is free.

BDo not assign the duplicate IPv6 address; use an address recovery path.

CMerge both IPv6 interfaces automatically under their duplicate address.

DDelegate duplicate-address response only to the transport layer.

Show answer

Answer: B DAD failure blocks assignment of that address; how a new identifier or manual recovery is obtained depends on the configuration mechanism.

Retrieval practice

Recall check 5 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q5A custom allocator makes 250 independent uniform choices from 65,536 local identifiers. The birthday approximation gives about a 37.8% chance of at least one duplicate. What is the strongest interpretation?

AExactly 37.8% of all packets lost in this fleet due to duplicate local identifiers.

BAbout a 37.8% chance of at least one duplicate in this random-choice domain.

CA hard limit of fewer than 250 devices for every 16-bit-address protocol.

DExactly 37.8% of the 65,536-value address space already occupied.

Show answer

Answer: B The result belongs to the stated independent, uniform, with-replacement model and says nothing by itself about a named protocol's capacity.

Retrieval practice

Recall check 6 of 6

Packet Pete says: answer from memory, then check your reasoning.

Q6A coordinator keeps a table of used values and always assigns an unused identifier. Why should an engineer not apply the independent birthday formula unchanged?

AAssignment without replacement prevents duplicate IDs while its table is correct.

BThe formula applies to decimal identifiers because decimal notation determines the collision probability for an identifier space.

CThe coordinator prevents every collision on the radio, not just duplicate IDs.

DThe collision probability becomes exactly 50% under coordinated allocation.

Show answer

Answer: A Coordination removes the independent-with-replacement assumption.

Print reference

Answers 1 of 2

Answer key.

A · Names, airtime, and shared retry timing need separate controls.
C · Each control targets a different collision mechanism.
B · Coordination or disjoint namespaces changes the assignment process so the servers cannot legitimately issue the same value.
B · DAD failure blocks assignment of that address; how a new identifier or manual recovery is obtained depends on the configuration mechanism.
B · The result belongs to the stated independent, uniform, with-replacement model and says nothing by itself about a named protocol's capacity.

Print reference

Answers 2 of 2

Answer key.

A · Coordination removes the independent-with-replacement assumption.