mirror of
https://github.com/zama-ai/concrete.git
synced 2026-02-15 07:05:09 -05:00
update code add results
This commit is contained in:
Ben
BIN
results/results_32_128.txt
Normal file
BIN
results/results_32_128.txt
Normal file
Binary file not shown.
BIN
results/results_64_128.txt
Normal file
BIN
results/results_64_128.txt
Normal file
Binary file not shown.
234
scripts.py
234
scripts.py
@@ -233,21 +233,68 @@ def automated_param_select_n(sd, n=None, q=2 ** 32, reduction_cost_model=BKZ.sie
|
||||
|
||||
|
||||
# initial estimate, to determine if we are above or below the target security level
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo, skip = {"bkw","dec","arora-gb","mitm"})
|
||||
security_level = get_security_level(estimate)
|
||||
z = inequality(security_level, target_security)
|
||||
|
||||
while z * security_level < z * target_security:
|
||||
print("estimating for n = {}, q, sd".format(n))
|
||||
try:
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo,
|
||||
skip={"bkw", "dec", "arora-gb", "mitm"})
|
||||
except Exception as e:
|
||||
print(e)
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo,
|
||||
skip={"bkw", "dec", "arora-gb", "mitm", "dual"})
|
||||
print("the estimate is {}".format(estimate))
|
||||
security_level = get_security_level(estimate)
|
||||
print("the security level is: {}".format(security_level))
|
||||
z = inequality(security_level, target_security)
|
||||
print("the result of Z is{}".format(z))
|
||||
|
||||
while z * security_level < z * target_security and n > 80:
|
||||
n += z * 8
|
||||
print("N = {}".format(n))
|
||||
print("SECURITY LEVEL = {}".format(security_level))
|
||||
alpha = sqrt(2 * pi) * sd / RR(q)
|
||||
print("estimating for n = {}, q, sd".format(n))
|
||||
try:
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo, skip = {"bkw","dec","arora-gb","mitm"})
|
||||
except:
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo,
|
||||
skip={"bkw", "dec", "arora-gb", "mitm", "dual"})
|
||||
security_level = get_security_level(estimate)
|
||||
|
||||
n += 1
|
||||
if (-1 * sd > log(q, 2)):
|
||||
print("target security level is unatainable")
|
||||
break
|
||||
|
||||
print("the finalised parameters are {}, {}, with a security level of {}".format(n, q, security_level))
|
||||
# final estimate (we went too far in the above loop)
|
||||
if security_level < target_security:
|
||||
n -= z * 8
|
||||
alpha = sqrt(2 * pi) * sd / RR(q)
|
||||
print("N = {}".format(n))
|
||||
print("SECURITY LEVEL = {}".format(security_level))
|
||||
try:
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo, skip = {"bkw","dec","arora-gb","mitm"})
|
||||
except:
|
||||
estimate = estimate_lwe(n, alpha, q, secret_distribution=secret_distribution,
|
||||
reduction_cost_model=reduction_cost_model, m=oo,
|
||||
skip={"bkw", "dec", "arora-gb", "mitm", "dual"})
|
||||
security_level = get_security_level(estimate)
|
||||
|
||||
print("the finalised parameters are n = {}, log2(sd) = {}, log2(q) = {}, with a security level of {}-bits".format(n,
|
||||
sd,
|
||||
log(q,
|
||||
2),
|
||||
security_level))
|
||||
|
||||
# final sanity check so we don't return insecure (or inf) parameters
|
||||
if security_level < target_security or security_level == oo:
|
||||
n = None
|
||||
|
||||
return n
|
||||
|
||||
return ZZ(n)
|
||||
|
||||
|
||||
def automated_param_select_sd(n, sd=None, q=2**32, reduction_cost_model=BKZ.sieve, secret_distribution=(0, 1),
|
||||
@@ -357,7 +404,7 @@ def generate_parameter_matrix(n_range, sd=None, q=2**32, reduction_cost_model=BK
|
||||
|
||||
sd_ = sd
|
||||
|
||||
for n in range(n_min, n_max):
|
||||
for n in range(n_min, n_max + 1):
|
||||
sd = automated_param_select_sd(n, sd=sd_, q=q, reduction_cost_model=reduction_cost_model,
|
||||
secret_distribution=secret_distribution, target_security=target_security)
|
||||
sd_ = sd
|
||||
@@ -366,6 +413,40 @@ def generate_parameter_matrix(n_range, sd=None, q=2**32, reduction_cost_model=BK
|
||||
return RESULTS
|
||||
|
||||
|
||||
def generate_parameter_matrix_sd(sd_range, n=None, q=2**32, reduction_cost_model=BKZ.sieve,
|
||||
secret_distribution=(0, 1), target_security=128):
|
||||
"""
|
||||
:param n_range: a tuple (n_min, n_max) giving the values of n for which to generate parameters
|
||||
:param sd: the standard deviation of the LWE error
|
||||
:param q: the LWE modulus (q = 2**32, 2**64 in TFHE)
|
||||
:param reduction_cost_model: the BKZ cost model considered, BKZ.sieve is default
|
||||
:param secret_distribution: the LWE secret distribution
|
||||
:param target_security: the target number of bits of security, 128 is default
|
||||
|
||||
TODO: we should probably parallelise this function for speed
|
||||
TODO: code seems to fail when the initial estimate is < target_security bits
|
||||
|
||||
EXAMPLE:
|
||||
sage: X = generate_parameter_matrix([788, 790])
|
||||
sage: X
|
||||
[(788, 4294967296, -20.0), (789, 4294967296, -20.0)]
|
||||
"""
|
||||
|
||||
RESULTS = []
|
||||
|
||||
# grab min and max value/s of n
|
||||
(sd_min, sd_max) = sd_range
|
||||
|
||||
n = n
|
||||
|
||||
for sd in range(sd_min, sd_max + 1):
|
||||
n = automated_param_select_n(sd, n=n, q=q, reduction_cost_model=reduction_cost_model,
|
||||
secret_distribution=secret_distribution, target_security=target_security)
|
||||
RESULTS.append((n, q, sd))
|
||||
|
||||
return RESULTS
|
||||
|
||||
|
||||
def generate_parameter_step(results, label = None, torus_sd = True):
|
||||
"""
|
||||
Plot results
|
||||
@@ -473,6 +554,7 @@ def test_params(n, q, sd, secret_distribution):
|
||||
|
||||
return est
|
||||
|
||||
|
||||
def generate_iso_lines(N = [256, 2048], SD = [0, 32], q = 2**32):
|
||||
|
||||
RESULTS = []
|
||||
@@ -491,6 +573,7 @@ def generate_iso_lines(N = [256, 2048], SD = [0, 32], q = 2**32):
|
||||
|
||||
return RESULTS
|
||||
|
||||
|
||||
def plot_iso_lines(results):
|
||||
|
||||
x1 = []
|
||||
@@ -529,10 +612,133 @@ def test_multiple_sd(n, q, secret_distribution, reduction_cost_model, split = 33
|
||||
return est, Y
|
||||
|
||||
|
||||
def estimate_lwe_sd(n, sd, q, secret_distribution, reduction_cost_model, skip = ("bkw","mitm","dec","arora-gb"), m = oo):
|
||||
def output_secret_distribution(m):
|
||||
"""
|
||||
generate the correct secret_distirbution for the given input
|
||||
:param m: the number of elements in the secret distribution
|
||||
"""
|
||||
|
||||
alpha = sqrt(2*pi) * sd/q
|
||||
x = estimate_lwe(n = n, alpha = alpha , q = q, m = m, secret_distribution = secret_distribution, reduction_cost_model = reduction_cost_model, skip = skip)
|
||||
# the code doesn't work for m < 2
|
||||
assert m >= 2
|
||||
|
||||
return x
|
||||
if m % 2 ==1:
|
||||
# m is odd
|
||||
b = (m - 1)/2
|
||||
secret_distribution = (-b, b)
|
||||
else:
|
||||
# m is even
|
||||
b = m / 2 - 1
|
||||
secret_distribution = (-b, b + 1)
|
||||
|
||||
return secret_distribution
|
||||
|
||||
def get_marcs_curves(n_range, q, m_max):
|
||||
|
||||
# the final result will be a list of m_max elements, each containing
|
||||
# a parameter matrix
|
||||
RESULTS = []
|
||||
|
||||
for m in range(2, m_max + 1):
|
||||
secret_distribution = output_secret_distribution(m)
|
||||
result_m = generate_parameter_matrix(n_range, sd=None, q=q, reduction_cost_model=BKZ.sieve,
|
||||
secret_distribution=secret_distribution, target_security=128)
|
||||
RESULTS.append(result_m)
|
||||
return RESULTS
|
||||
|
||||
|
||||
def get_marcs_curves_n(sd_range, q, m_max):
|
||||
|
||||
# the final result will be a list of m_max elements, each containing
|
||||
# a parameter matrix
|
||||
RESULTS = []
|
||||
|
||||
for m in range(2, m_max + 1):
|
||||
secret_distribution = output_secret_distribution(m)
|
||||
result_m = generate_parameter_matrix_sd(n = None, sd_range=sd_range, q=q, reduction_cost_model=BKZ.sieve,
|
||||
secret_distribution=secret_distribution, target_security=128)
|
||||
RESULTS.append(result_m)
|
||||
return RESULTS
|
||||
|
||||
|
||||
def tabulate_results(results):
|
||||
""" Put the results from get_marcs_curves into a LaTeX table
|
||||
"""
|
||||
|
||||
new_results = []
|
||||
num_results = len(results[0])
|
||||
num_entries = len(results)
|
||||
|
||||
key = []
|
||||
key.append("n")
|
||||
key.append("q")
|
||||
for i in range(num_entries):
|
||||
key.append("m = {}".format(i + 2))
|
||||
|
||||
new_results.append(key)
|
||||
|
||||
|
||||
|
||||
|
||||
for j in range(num_results):
|
||||
result_j = []
|
||||
result_j.append(results[0][j][0])
|
||||
result_j.append(int(log(results[0][0][1],2)))
|
||||
|
||||
for i in range(num_entries):
|
||||
result_j.append(int(results[i][j][2]))
|
||||
|
||||
new_results.append(result_j)
|
||||
|
||||
return new_results
|
||||
|
||||
def tabulate_results_sd(results):
|
||||
|
||||
new_results = []
|
||||
num_results = len(results[0])
|
||||
num_entries = len(results)
|
||||
|
||||
key = []
|
||||
key.append("sd")
|
||||
key.append("q")
|
||||
|
||||
for i in range(num_entries):
|
||||
key.append("m = {}".format(i + 2))
|
||||
|
||||
new_results.append(key)
|
||||
|
||||
for j in range(num_results):
|
||||
result_j = []
|
||||
result_j.append(results[0][j][2])
|
||||
result_j.append(int(log(results[0][0][1],2)))
|
||||
|
||||
for i in range(num_entries):
|
||||
try:
|
||||
result_j.append(int(results[i][j][0]))
|
||||
except:
|
||||
result_j.append(str(results[i][j][0]))
|
||||
|
||||
new_results.append(result_j)
|
||||
|
||||
return new_results
|
||||
|
||||
# code to cross-check the security levels for marc/pascal results
|
||||
# sage: with open("results_32_128.txt", "rb") as fp: # Unpickling
|
||||
# ....: ... X = pickle.load(fp)
|
||||
# res = []
|
||||
# sage: for i in range(len(X)):
|
||||
# ....: x = X[i]
|
||||
# ....: m = i + 2
|
||||
# ....: secret_distribution = output_secret_distribution(m)
|
||||
# ....: for (n, q, sd) in x:
|
||||
# ....: if n is not None:
|
||||
# ....: sd = 2**(sd)
|
||||
# ....: alpha = sqrt(2*pi) * sd
|
||||
# ....: print((n, q, sd))
|
||||
# ....: try:
|
||||
# ....: _ = estimate_lwe(n, alpha, q, secret_distribution = secret_distribution, reduction_cost_model = BKZ.sieve, skip = ("arora-gb", "mitm", "bkw", "dec"))
|
||||
# ....: except:
|
||||
# ....: _ = estimate_lwe(n, alpha, q, secret_distribution = secret_distribution, reduction_cost_model = BKZ.sieve, skip = ("arora-gb", "mitm", "bkw", "dec", "dual"))
|
||||
# ....: else:
|
||||
# ....: print("None")
|
||||
# ....: res.append(get_security_level(_))
|
||||
# ....:print(min(res))
|
||||
Reference in New Issue
Block a user