def main(): #define two players, chose gameType as the following: 1-implementation 1 (theoretical) ; 2 - implementation 2(practically form 1); 3 - implementation 3(practically form 2) - recommended player1 = Player(10001, 25, 2, 1700.00, 1, game_Type=3) player2 = Player(10001, 16, 17, 2500.00, 2, game_Type=3) #set links between players. (They have to communicate one with each other) player1.setNeighborPlayer(player2) player2.setNeighborPlayer(player1) #update gains #player1.measureGains() #player2.measureGains() player1.setDirectGain(GainCalculations.getAverageGain(player1.coordinator_id, player1.tx_node.node_id, player1.rx_node.node_id, year=2013, month=8, day=23)) player1.setCrossGain(GainCalculations.getAverageGain(player1.coordinator_id, player2.tx_node.node_id, player1.rx_node.node_id, year=2013, month=8, day=23)) player2.setDirectGain(GainCalculations.getAverageGain(player2.coordinator_id, player2.tx_node.node_id, player2.rx_node.node_id, year=2013, month=8, day=23)) player2.setCrossGain(GainCalculations.getAverageGain(player2.coordinator_id, player1.tx_node.node_id, player2.rx_node.node_id, year=2013, month=8, day=23)) #if you want to see some info about players player1.printPlayerInfo() player2.printPlayerInfo() #check convergence. See if the topology suits the game conditions if not checkConvergence(player1.direct_gain, player2.cross_gain, player1.cross_gain, player2.direct_gain): return livePlot = gameLivePlot(player1, player2) #start live plotting livePlot.startPloting() #start Power Allocation threads player1.startGame() player2.startGame() time.sleep(3) #trigger an event by changing the power of one player player1.unbalance() #wait for the threads to finish while player1.powerAllocation.isAlive() or player2.powerAllocation.isAlive() or livePlot.is_alive(): time.sleep(3) #if you want to plot the results iterations.plotIterations(player1.player_number, player1.tx_node.node_id, player1.rx_node.node_id, player2.player_number, player2.tx_node.node_id, player2.rx_node.node_id, int(player1.cost), int(player2.cost), TruncatedValues=False, saveImage=True) choice = raw_input("Do you want to start PlayerApp threads(sends packets continuously) - yes/no:") if choice.lower() == "yes": player1.playerApp.start() player2.playerApp.start()
def getAFewParameters(): #From the measured gain, you can get a few results #If you just want to get a list with measured gains: print GainCalculations.getFileResults(10001, 25, 2) #You will find informations about: calculated gain, received RSSI, Noise power, transmitted power, date #if you just want the average gain: print GainCalculations.getAverageGain(10001, 25, 2) print "%.3f dB" %(10.00*math.log10(GainCalculations.getAverageGain(10001, 25, 2))) #if you just want to find info about the noise power GainCalculations.getAverageNoise(10001, 25, 2) GainCalculations.getMinMaxNoise(10001, 25, 2)
def bestResponseAsAfunctionOfCostAndPrx(): #plot best response variation as function of I+N and cost player = Player(10001, 25, 2, 1000.00, player_number=1, game_Type=0) #give the node id which cause interference to player tx2_node_id = 16 #I want to determine the maximum level of interference #average direct gain hii = GainCalculations.getAverageGain(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23) #maximum cross gain hji = GainCalculations.getMinMaxGain(10001, tx2_node_id, player.rx_node.node_id, year=2013, month=8, day=24) #get average noise noise = GainCalculations.getAverageNoise(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23) max_interference_and_noise = 0.001*hji[1] + noise max_interference_and_noise = 10.00*math.log10(max_interference_and_noise/0.001) interference_and_noise = numpy.arange(max_interference_and_noise, max_interference_and_noise-10, -2) cost = numpy.arange(100, 10000, 0.1) #now plot results plot.ioff() plot.clf() plot.grid() plot.title("B%d (c%d, I+N)" %(player.player_number, player.player_number)) plot.xlabel("c%d" %(player.player_number)) plot.ylabel("B%d [dBm]" %(player.player_number)) for i in interference_and_noise: tmp_list = [] tmp_cost = [] for c in cost: tmp_bi = getBi(c, math.pow(10.00, i/10.00)*0.001, 10.00*math.log10(hii)) if tmp_bi!=None: tmp_list.append(tmp_bi) tmp_cost.append(c) plot.plot(tmp_cost, tmp_list, label = "I+N=%.1f dBm" %i) plot.plot([],[],label = "h%d%d = %.3f dB" %(player.player_number, player.player_number, 10.00*math.log10(hii))) plot.axhspan(-55, 0, alpha = 0.1) plot.legend(bbox_to_anchor=(1.05, 1.05)) plot.show() #bestResponseAsAfunctionOfCostAndPrx() #bestResponseAsAFunctionOfDirectGain() #bestResponseAsAFunctionOfRx()
def bestResponseAsAFunctionOfRx(): player = Player(10001, 25, 2, 1000.00, 1) #get gain measurements until 23 august player.setDirectGain(GainCalculations.getAverageGain(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23)) #find max Prx_dBm in which Bi > -55 dBm Prx_dBm = -100 while True: tmp_bi = getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, 10.00*math.log10(player.direct_gain)) if tmp_bi is None or tmp_bi<-55: Prx_dBm-=0.001 break Prx_dBm+=0.0001 #now plot results plot.ioff() plot.clf() plot.grid() plot.title("B%d=f(I+N)" %(player.player_number)) plot.xlabel("I+N [dBm]") plot.ylabel("B%d [dBm]" %(player.player_number)) Prx_crt = Prx_dBm Prx_inf = Prx_dBm-20 Prx_step = 0.001 Bi = [] Prx = [] while Prx_crt>=Prx_inf: best_response = getBi(player.cost, math.pow(10.00, Prx_crt/10.00)*0.001, 10.00*math.log10(player.direct_gain)) Bi.append(best_response) Prx.append(Prx_crt) Prx_crt-=Prx_step plot.plot(Prx, Bi, label = "h%d%d=%.1f dBm" %(player.player_number,player.player_number,10.00*math.log10(player.direct_gain)), linewidth = 2) plot.axis([min(Prx)-1, max(Prx)+1, min(Bi)-1, max(Bi)+1]) leg = plot.legend(loc = 'upper right', fontsize = 18) leg.get_frame().set_alpha(0.6) plot.show()
def bestResponseAsAFunctionOfDirectGain(): #formula to test: Bi = (1/c) - (Prx/hii), where c is constant, Prx= ct #define a player player = Player(10001, 25, 2, 1000.00, 1) #player = Player(10001, 16, 17, 1000.00, 2) #get min and max channel gains measured until 23 august tmp = GainCalculations.getMinMaxGain(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23) #tmp = [min linear gain, maximum linear gain] min_hii = 10.00*math.log10(tmp[0]) max_hii = 10.00*math.log10(tmp[1]) #define a hii array hii = numpy.arange(min_hii, max_hii, 0.05) #find max Prx_dBm in which Bi > -55 dBm Prx_dBm = -100 while True: tmp_bi = getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, min_hii) if tmp_bi is None or tmp_bi<-55: #return to the previous Prx_dBm Prx_dBm-=0.0001 break Prx_dBm+=0.0001 average_gain = GainCalculations.getAverageGain(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23) standard_deviation = GainCalculations.getStandardDeviation(player.coordinator_id, player.tx_node.node_id, player.rx_node.node_id, year=2013, month=8, day=23) #now plot results plot.ioff() plot.clf() plot.grid() plot.title("B%d=f(h%d%d, I+N)" %(player.player_number, player.player_number, player.player_number)) plot.xlabel("h%d%d [dB]" %(player.player_number, player.player_number)) plot.ylabel("B%d [dBm]" %(player.player_number)) #change rc param plot.rcParams.update({'font.size': 20}) #I just want to add a text to legend plot.plot([0], [0], alpha = 0, label = "c%d=%d" %(player.player_number, player.cost)) #now plot bi Prx_crt = Prx_dBm Prx_step = 3 Prx_inf = Prx_crt - 10 max_bi = -float("inf") min_bi = float("inf") markers = [".", "*", "+", "h", "x", "_"] marker_index = 0 while Prx_crt >= Prx_inf: Bi = [] for i in hii: best_response = getBi(player.cost, math.pow(10.00, Prx_crt/10.00)*0.001, i) Bi.append(best_response) if best_response > max_bi: max_bi = best_response if best_response < min_bi: min_bi = best_response plot.plot(hii, Bi, linewidth = 2.5, label = "Prx %.2f dBm" %(Prx_crt), marker = markers[marker_index], markersize = 5) if marker_index<len(markers): marker_index+=1 else: marker_index = 0 Prx_crt-=Prx_step #set axis limits plot.axis([min(hii)-0.5, max(hii)+0.5, min_bi-2, max_bi+2]) #set ticks plot.xticks(numpy.arange(min(hii),max(hii), 2)) plot.yticks(numpy.arange(min_bi, max_bi, 2)) #plot a vertical line with the average gain plot.vlines(10.00*math.log10(average_gain), plot.axis()[2], plot.axis()[3], "red", label = "Mean gain=%.2f dB" %(10.00*math.log10(average_gain)), linestyle = "--", linewidth = 2) #plot vertical lines with the +- standard deviation max_std_hii = 10.00*math.log10(average_gain+standard_deviation) min_std_hii = 10.00*math.log10(average_gain-standard_deviation) #plot.vlines(min_std_hii, plot.axis()[2], plot.axis()[3], "black", linestyle = "--", linewidth = 1) #plot.vlines(max_std_hii, plot.axis()[2], plot.axis()[3], "black", linestyle = "--", linewidth = 1) #plot some spans for +- standard deviation plot.axvspan(min_std_hii, max_std_hii, facecolor = "gray", alpha = 0.2) #Plot horizontal arrows #plot arrows for +- standard deviation arrow_length = math.fabs(max_std_hii - min_std_hii) plot.arrow(min_std_hii, min_bi + 1, arrow_length, 0, head_width = 0.02*(math.fabs(max(hii)-min(hii))), length_includes_head = True, color = "green",head_length = 0.02*(math.fabs(max(hii)-min(hii))), linewidth = 1.0) plot.arrow(max_std_hii, min_bi + 1, -arrow_length, 0, head_width = 0.02*(math.fabs(max(hii)-min(hii))), length_includes_head = True, color = "green",head_length = 0.02*(math.fabs(max(hii)-min(hii))), linewidth = 1.0) plot.text(min_std_hii+arrow_length/3, min_bi +1.1, "%.3f dB" %arrow_length, fontsize = 18, color = "green") #plot horizontal lines for best response variation #plot.hlines(getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, max_std_hii), min(hii), max(hii), color = "black", linestyle = "--", linewidth = 1) #plot.hlines(getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, min_std_hii), min(hii), max(hii), color = "black", linestyle = "--", linewidth = 1) #plot.axhspan(getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, min_std_hii), getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, max_std_hii), facecolor = "blue", alpha = 0.1) #plot arrows for best response variation #max_std_Bi = getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, max_std_hii) #min_std_Bi = getBi(player.cost, math.pow(10.00, Prx_dBm/10.00)*0.001, min_std_hii) #arrow_length = math.fabs(max_std_Bi - min_std_Bi) #plot.arrow(min(hii) +0.5, max_std_Bi, 0, -arrow_length, head_width = 0.01*(math.fabs(max(hii)-min(hii))), length_includes_head = True, color = "blue", head_length = 0.01*(math.fabs(max(hii)-min(hii))), linewidth = 1.3) #plot.arrow(min(hii) +0.5, min_std_Bi, 0, +arrow_length, head_width = 0.01*(math.fabs(max(hii)-min(hii))), length_includes_head = True, color = "blue", head_length = 0.01*(math.fabs(max(hii)-min(hii))), linewidth = 1.3) #plot.text(min(hii) +0.5, max_std_Bi, "%.3f dBm" %arrow_length, fontsize = 18, weight = 500, color = "blue") leg = plot.legend(loc = "center right", fontsize = 18, bbox_to_anchor=(1, 0.5)) leg.get_frame().set_alpha(0.6) #maximize the window mng = plot.get_current_fig_manager() mng.resize(*mng.window.maxsize()) fig = plot.gcf() fig.set_size_inches( (19, 11) ) #plot.savefig("/home/ciprian/Pictures/best response/%s%d.jpg" %("bi_f(hii)_player", player.player_number), dpi=250) plot.show()