def SDPTDoALocate(self, RN1, RN2, TDoA, TDoAStd):
"""
Apply SDP approximation and localization
"""
RN1 = cvxm.matrix(RN1)
RN2 = cvxm.matrix(RN2)
TDoA = cvxm.matrix(TDoA)
c = 3e08
RDoA = c*TDoA
RDoAStd=cvxm.matrix(c*TDoAStd)
mtdoa,ntdoa=cvxm.size(RN1)
Im = cvxm.eye(mtdoa)
Y=cvxm.optvar('Y',mtdoa+1,mtdoa+1)
t=cvxm.optvar('t',ntdoa,1)
prob=cvxm.problem(cvxm.minimize(cvxm.norm2(t)))
prob.constr.append(Y>=0)
prob.constr.append(Y[mtdoa,mtdoa]==1)
for i in range(ntdoa):
X0=cvxm.matrix([[Im, -cvxm.transpose(RN1[:,i])],[-RN1[:,i], cvxm.transpose(RN1[:,i])*RN1[:,i]]])
X1=cvxm.matrix([[Im, -cvxm.transpose(RN2[:,i])],[-RN2[:,i], cvxm.transpose(RN2[:,i])*RN2[:,i]]])
prob.constr.append(-RDoAStd[i,0]*t[i]<cvxm.trace(X0*Y)+cvxm.trace(X1*Y)-RDoA[i,0]**2)
prob.constr.append(RDoAStd[i,0]*t[i]>cvxm.trace(X0*Y)+cvxm.trace(X1*Y)-RDoA[i,0]**2)
prob.solve()
Pval=Y.value
X_cvx=Pval[:2,-1]
return X_cvx
def SDPRSSLocate(self, RN, PL0, d0, RSS, RSSnp, RSSStd, Rest):
RoA=self.getRange(RN, PL0, d0, RSS, RSSnp, RSSStd, Rest)
RN=cvxm.matrix(RN)
RSS=cvxm.matrix(RSS)
RSSnp=cvxm.matrix(RSSnp)
RSSStd=cvxm.matrix(RSSStd)
PL0=cvxm.matrix(PL0)
RoA=cvxm.matrix(RoA)
mrss,nrss=cvxm.size(RN)
Si = array([(1/d0**2)*10**((RSS[0,0]-PL0[0,0])/(5.0*RSSnp[0,0])),(1/d0**2)*10**((RSS[1,0]-PL0[1,0])/(5.0*RSSnp[1,0])),(1/d0**2)*10**((RSS[2,0]-PL0[2,0])/(5.0*RSSnp[2,0])),(1/d0**2)*10**((RSS[3,0]-PL0[3,0])/(5.0*RSSnp[3,0]))])
#Si = array([(1/d0**2)*10**(-(RSS[0,0]-PL0[0,0])/(5.0*RSSnp[0,0])),(1/d0**2)*10**(-(RSS[0,1]-PL0[1,0])/(5.0*RSSnp[0,1])),(1/d0**2)*10**(-(RSS[0,2]-PL0[2,0])/(5.0*RSSnp[0,2])),(1/d0**2)*10**(-(RSS[0,3]-PL0[3,0])/(5.0*RSSnp[0,3]))])
Im = cvxm.eye(mrss)
Y=cvxm.optvar('Y',mrss+1,mrss+1)
t=cvxm.optvar('t',nrss,1)
prob=cvxm.problem(cvxm.minimize(cvxm.norm2(t)))
prob.constr.append(Y>=0)
prob.constr.append(Y[mrss,mrss]==1)
for i in range(nrss):
X0=cvxm.matrix([[Im, -cvxm.transpose(RN[:,i])],[-RN[:,i], cvxm.transpose(RN[:,i])*RN[:,i]]])
prob.constr.append(-RSSStd[i,0]*t[i]<Si[i]*cvxm.trace(X0*Y)-1)
prob.constr.append(RSSStd[i,0]*t[i]>Si[i]*cvxm.trace(X0*Y)-1)
prob.solve()
Pval=Y.value
X_cvx=Pval[:2,-1]
return X_cvx
def SDPToALocate(self, RN, ToA, ToAStd):
"""
Apply SDP approximation and localization
"""
RN = cvxm.matrix(RN)
ToA = cvxm.matrix(ToA)
c = 3e08 # Speed of light
RoA = c*ToA
RoAStd = c*ToAStd
RoAStd = cvxm.matrix(RoAStd)
mtoa,ntoa=cvxm.size(RN)
Im = cvxm.eye(mtoa)
Y=cvxm.optvar('Y',mtoa+1,mtoa+1)
t=cvxm.optvar('t',ntoa,1)
prob=cvxm.problem(cvxm.minimize(cvxm.norm2(t)))
prob.constr.append(Y>=0)
prob.constr.append(Y[mtoa,mtoa]==1)
for i in range(ntoa):
X0=cvxm.matrix([[Im, -cvxm.transpose(RN[:,i])],[-RN[:,i], cvxm.transpose(RN[:,i])*RN[:,i]]])
prob.constr.append(-t[i]<(cvxm.trace(X0*Y)-RoA[i]**2)*(1/RoAStd[i]))
prob.constr.append(t[i]>(cvxm.trace(X0*Y)-RoA[i]**2)*(1/RoAStd[i]))
prob.solve()
Pval=Y.value
X_cvx=Pval[:2,-1]
return X_cvx
def get_energy_per_quanto_state_all(self, baseFileName, convexOpt=False):
"""This function evaluates the .pwr file and calculates the individual
power consumption per state for every node. It then sets the variable
statePower on every node to a dictionary, where the keys are the
string representation of the state, and the value is the average power
consumption for that state.
This function expects a very specific .pwr file, where one line starts
with "#states:". This line encodes all the states that are considered
by quanto.
"""
for n in self.nodes:
f = open("%s.%s.log.pwr"%(baseFileName, n.ip), "r")
X = []
Y = []
W = []
totalTime = 0
totalEnergy = 0
states = []
maxEntries = 0
for line in f:
l = line.strip().split()
if len(l) > 0 and l[0] == "#states:":
# this line encodes the names of all the states.
states = l[1:]
continue
if len(states) == 0 or len(l) != len(states)+3:
# +2 comes from the icount and time field
continue
#time is in uS, convert it to seconds
time = float(l[-3])/1e6
icount = int(l[-2])
occurences = int(l[-1])
# cut away the time and icount values
l = l[0:-3]
activeStates = []
for s in l:
if s == '-':
s = '0'
#continue
activeStates.append(int(s))
# add the constant power state
activeStates.append(1)
if len(activeStates) > maxEntries:
maxEntries = len(activeStates)
if time <= 0 or icount <= 0:
# FIXME: this is a wrong line at the end of the quanto files. I
# don't know why this happens!!!
continue
E = n.get_power(icount, time)
if E == -1:
raise CalibrationError, "Node with IP %s is not calibrated! \
Did you forget to load the calibration file?"%(n.ip,)
if E < 0:
raise CalibrationError, "Node with IP %s returned a \
negative Energy value %f for icount %d, time %f!"%(n.ip, E, icount, time)
continue
X.append(activeStates)
Y.append(E)
W.append(numpy.sqrt(E*time))
totalTime += time
totalEnergy += E*time
# filter out the incomplete datasets
Xnew = []
Ynew = []
Wnew = []
for i in range(len(Y)):
if len(X[i]) == maxEntries:
Xnew.append(X[i])
Ynew.append(Y[i])
Wnew.append(W[i])
X = numpy.matrix(Xnew)
Y = numpy.matrix(Ynew)
W = numpy.matrix(numpy.diag(Wnew))
# filter states with all 0's
states.append('const')
deletedLines = 0
deletedStates = []
alwaysOnStates = []
# iterate through all the states, except the 'const' state
for i in range(len(states)-1):
correctedI = i - deletedLines
if numpy.sum(X.T[correctedI]) == 0:
deletedStates.append(states[correctedI])
X = numpy.delete(X, numpy.s_[correctedI:correctedI+1], axis=1)
states = numpy.delete(states, correctedI)
deletedLines += 1
elif numpy.sum(X.T[correctedI]) == len(X):
# this state is always active. W have to remove them and
# put them into the "const" category!
alwaysOnStates.append(states[correctedI])
X = numpy.delete(X, numpy.s_[correctedI:correctedI+1], axis=1)
states = numpy.delete(states, correctedI)
deletedLines += 1
# search for linear dependent lines
#for i in range(len(X)):
# #print X[i]
# for j in range(i+1, len(X)):
# #if numpy.sum(X[i]) == numpy.sum(X[j]):
# # print X[i], X[j]
# equal = True
# for m in range(X[i].shape[1]):
# if X[i,m] != X[j,m]:
# equal = False
# break
# if equal:
# print X[i], X[j]
# maxEntries includes the const state, which is not in the states
# variable yet
#states = states[:maxEntries - 1]
#xtwx = X.T*X
#for i in range(len(xtwx)):
# print xtwx[i]
#print states
#(x, resids, rank, s) = numpy.linalg.lstsq(W*X, W*Y.T)
#(x, resids, rank, s) = numpy.linalg.lstsq(X, Y.T)
if cvxAvailable and convexOpt:
A = cvxmod.matrix(W*X)
b = cvxmod.matrix(W*Y.T)
x = cvxmod.optvar('x', cvxmod.size(A)[1])
print A
print b
p = cvxmod.problem(cvxmod.minimize(norm2(A*x - b)), [x >= 0])
#p.constr.append(x |In| probsimp(5))
p.solve()
print "Optimal problem value is %.4f." % p.value
cvxmod.printval(x)
x = x.value
else:
try:
x = numpy.linalg.inv(X.T*W*X)*X.T*W*Y.T
except numpy.linalg.LinAlgError, e:
sys.stderr.write("State Matrix X for node with IP %s is singular. We did not \
collect enough energy and state information. Please run the application for \
longer!\n"%(n.ip,))
sys.stderr.write(repr(e))
sys.stderr.write("\n")
sys.stderr.flush()
n.statePower = {}
n.alwaysOffStates = []
n.alwaysOnStates = []
continue
n.statePower = {}
for i in range(len(states)):
# the entries in x are matrices. convert them back into a
# number
n.statePower[states[i]] = float(x[i])
n.alwaysOffStates = deletedStates
n.alwaysOnStates = alwaysOnStates
n.averagePower = totalEnergy / totalTime
