The Tangle network, a product of IOTA, promises global scale micropayments with low fees designed for the Internet of Things. Can this next generation distributed ledger technology work on the IoT's edge? Let's take a look.
Evolving the trustless distributed blockchain pioneered by Bitcoin, the Tangle ledger is a blockchainless system designed to operate on the IoT's edge - on the devices themselves. These Tangle-enabled devices could, for example, purchase access to wireless networks on demand using "IOTA coin". Unlike Bitcoin, which relies on massively expensive "proofs of work" (i.e. hashing power) to authenticate and maintain the distributed blockchain ledger, the Tangle ledger is specified by the network topology itself in the form of a DAG - Directed Acyclic Graph. Nodes may generate a transaction but there is a catch - for the transaction to be valid it must approve two other transactions (presumably incoming toward the node).
The Tangle ledger protocol specifies that: "For the transaction to be valid, the node must solve a cryptographic puzzle (which may be computationally demanding) similar to those in the Bitcoin mining" (Quoted from the IOTA Whitepaper). This statement raises concerns as to the viability of Tangle for IoT microtransactions. The requirement to perform something computationally demanding, and hence power consuming, works against IoT's fundamental principles of being largely made up of lower-power and in many cases battery-operated devices.
The Tangle network behavior is complex. As opposed to Bitcoin's relative simplicity, a number of complex equations serve to define the Tangle protocol. Complexity is not necessarily a good thing. As well, the complex attack scenarios analyzed in the paper offer the reader no level of comfort as there surely are many more possible attacks (see this recent IOTA blog post).
For IOTA's Tangle to be viable as an IoT cryptocurrency, it must not rely on solving puzzles that consume energy.
It remains to be seen exactly how the network remains stable and resistant to attacks over time, or why it must be necessarily be "public". A good portion of the IoT is already controlled by access points with security requirements and keys that must be provisioned on the devices before access; these keys are managed by the access point provider. For example, a LoRaWAN operator can require that the devices accessing its network must be pre-registered to obtain an identity and AES key to secure the communication link. In this context, then, a more purpose-built transaction system with a central trust authority may make more sense.
The IOTA Tangle is an interesting concept and more than anything serves to further the discussion about what distributed micropayment system might look like in a world of connected things. If we get a chance to work with a real Tangle network on our connected things we'll post updates.
Update: 9/23/18
IOTA has failed to demonstrate any use for IoT. Burdening the battery-powered devices themselves with the responsibility for reaching global state consensus hardly seems like the right way to allocate resources. Plus, blockchain for IoT is about security, and IOTA has gone off the rails there, making up their own security schemes. Trust in a globally recognized and tried and true security scheme like blockchain's ECDSA (used in Bitcoin, Ethereum, EOS, countless others) is invaluable. RIP IOTA.
Instead, look at interesting IoT-blockchain projects around EOS:
https://github.com/firmwaremodules/Contiki-EOS
https://github.com/eosiot
https://eosiot.io