# Larkworks

# Meeting notes 07-14-2013

- Kyber [engineering plan](https://hackmd.io/@alxiong/kyber-plan)
- send invite for July 26th mid term report to EF
- targets:
- Underly math
- lattice operations
- Hash functions
- SIS based
- Vector Commitment
- Merkle tree in Chipmunk
- Kyber scheme
- Chipmunk scheme
- Alex will look at ZZ's PR and tell ZZ what part he is working on
- Marcin will work on the Ring implementation
- ZZ will port chipmunk code into Larkworks
- Have a meeting on Tuesday July 25th

# Meeting notes 06-23-2023

- done
- zhenfei
- implement ff's Field/PrimeField for u64
- implement FFTs for small domains (WIP)

- sprint
- zhenfei
- implement Distributor struct (allows for gaussian/uniform/binomial)
- implement Lattice strcut (only have additions now, may add more functions later)

# Meeting notes 05-12-2023

- marcin
- sage code [PR](https://github.com/larkworks/larkworks/pull/1/files)

- zhenfei
- rust code base for polynomial operations
- used this code base for Chipmunk paper

- Alex
- EC sharing [notes](https://gist.github.com/alxiong/8d431224bc54d93901de373e1cf518a5)

# Meeting notes 04-23-2023

- code base:
- link: https://github.com/larkworks/larkworks
- added polynomial module from ZZ's previous work
- funding updates:
- Got an approval from EF
- No action needed for now
- misc:
- meet biweekly on Friday morning 9 am ET.
- Progress update:
- Alex:
- spent some time reading [Decade in Lattice [Pei15]](https://eprint.iacr.org/2015/939) (currently finishing until Chapter 5.2.1 on CPA-secure scheme, precursors of Kyber's CPA-part), Micciancio's [lec notes](https://cseweb.ucsd.edu/classes/sp14/cse206A-a/) on some lattic basic and [harmonic analysis](https://cseweb.ucsd.edu/classes/fa21/cse206A-a/LecGaussian.pdf), helps me a great deal with necessary foundation.
- Apologize that I haven't start implementing Kyber yet. I'm sure I will read Kyber spec more carefully and start my toy impl next week.
- I'll be on CET for a month.

todo:
- ZZ: change the modulus of existing code to kyber modulus
- ZZ: convert the polynomial code base ring arithmetics
- Alex: present kyber overview as a hackmd doc
- Marcin: finish sage code and integrate the test code
- in the same repo as larkworks

# Meeting notes 04-07-2023
## Scope/proposal

A library for lattice based cryptography, may include

- basic polynomial and matrix arithmetics
- various samplers such as GPV, MP12, Falcon
- (optional) lwe/sis hardness estimator
- PKI primitives: DualRegev, kyber, dilihtium, NTRU, Falcon
- hash fucntions: Swifft
- commitment schemes: BDLOP
- verifiable random fucntions
- lattice based SNARKs.

### Phase 1

- basic polynomial and matrix arithmetics (6 weeks)
- accompanied with sage scripts for validity tests
- allows for customized ring and field setting
- implement butterfly FFTs
- specs: NIST submission

- various samplers such as Falcon (6 weeks)

- (optional) lwe/sis hardness estimator
- integrate lwe-estimator python script

- PKI primitives: kyber (2 weeks), dilihtium (2 weeks), Falcon (2 weeks)
- performance comparisons with NIST submissions
- passes known answer tests (KAT) from NIST's specs

### Phase 2

- hash fucntions: Swifft
- commitment schemes: BDLOP

### Phase 3
- Plonk with BDLOP commitment
- Plonk with FRI + Swifft hash

### Bonus (not essential for ZKP but would be nice to have)
- Squrriel with Larkworks backend
- PKI primitives: DualRegev, NTRU
- various samplers such as GPV, MP12

do we want to allow for FHEs, if so we need
- RNS
- large modulus

todos:
- larkwork monorepo renaming
- ship some of existing code from ZZ
- specs kyber
- how to define polynomial trait (draft in 2 weeks)
- build a PoC of kyber (in 4 weeks)
- biweekly meeting same time Friday morning 9am ET
- prepare the grant application and submit it

- next meeting
- chat about kyber spec
- draft implementation of PrimeField, and perhaps a different definition of Polynomials

# Spec

## Arithmetic

- [x] Field
- As defined in ff-0.13
- [x] PrimeField: Field
- As defined in ff-0.13
- [ ] Polynomial<F:Field>
- [x] First iteration https://github.com/larkworks/larkworks/blob/45178f7f186c4cf62a71186a57f8a6f70615e523/polynomial/src/lib.rs#L31
- Move samplers out to Samplers etc.
- code need refactoring
- add school book muls

- [x] SparsePolynomial<F:Field>
- From<Polynomial> and To<Polynomial>
- [ ] Vector (Alex)
- NTTVector: Vector<F: PrimeField>
- convertsion from/to polynomials via NTTs
- LatticeVector: Vector<F: PrimeField> (supporting generic lattices; low priority)
- convertsion from/to polynomials via casting
- F does not need to be NTT friendly
- Do we need a seperate abstraction for prime field that is not NTT friendly?
- [ ] Matrix<E>
- E == Field => generic lattices
- E == Ring => modular lattices
- [ ] Ring(M)
- Move addtion/multiplications to Ring
- Define a new additon/multiplication over Z (seems not useful, low priority)
- wrapper of polynomial, with addion/multiplications over the ring
- [ ] Lattice
- what are essential interfaces here?
- instanciation is a wrapper of Matrix<Field>
- [ ] IdealLattice: Lattice
- instanciation is a wrapper of Ring
- [ ] ModularLattice
- instanciation is a wrapper of Matrix<R>
- [ ] Distribution
- hard code for Gaussian/uniform for now?
- [ ] Sampler
- a trait bound for Ring, Lattice, Vector?