Updated index.rst

index.rst

@@ -734,7 +734,7 @@ Drones have distinct software and hardware for mission plan (route map), flight
 	23: built spaces, non-residential, 6m < building height <= 15m
 	24: built spaces, non-residential, 15m < building height <= 30m
 	25: built spaces, non-residential, building height > 30m. 
-Segregating Non-residential as Commercial or Central Business District (CBD) and Industrial (Manufacturing-IT-ITES) splits earlier 3-color classification into desired 4-coloring and there are kernel density based clustering solutions suggested for identifying CBD : 1) Building an Urban Spatial Structure from Urban Land Use Data: An Example Using Automated Recognition of the City Centre - https://www.mdpi.com/2220-9964/6/4/122/htm 2) Delimitating Urban Commercial Central Districts by Combining Kernel Density Estimation and Road Intersections: A Case Study in Nanjing City, China - https://www.mdpi.com/2220-9964/8/2/93/htm . Moran's I and Gini Index are measures of dispersion and unequal lopsided distribution of growth within an urban sprawl. Graph connecting nightlights circular vertices by edges, could be an aerial transportation graph for autonomous delivery Drones between cities. Electricity Consumption and NightLights are mostly limited to Residential,Commercial and Manufacturing-IT-ITES regions of an urban sprawl and Agricultural land-Greenery-Waterbodies are excluded because of almost nil night lighting. Thus only 3 of the 4 colors of an Urban sprawl FaceGraph are covered by NightLights urban sprawl estimation excluding Farm lands which could be an error-free guideline for Urban planning and Expansion. There is an unusual theoretical aspect to NASA VIIRS NightLights urban area circles - Circular Urban areas of varied radii are packed (close or loose) on 2 dimensional space of a nation which makes it a variant of Space filling or Unequal Circle packing problem. Some bustling urban areas in VIIRS imagery of India 2017 are not heavily lit which could be inferred as "NightLights imply urbanization while converse may not be true (or) NightLights sprawl lowerbounds an Urban sprawl". NeuronRain implements ranking of urban sprawls based on areas of segmented Contour polynomials. Circles or contour polynomials bounding urban areas unequally fill the 2D space and ratio of sum of urban contour areas to area of the country is the urbanization percentage metric. Set of Urban area contours could be grouped by unsupervised clustering algorithms (e.g Neural Networks, DBSCAN, Voronoi Tessellation) using euclidean distances amongst their centroids from which juxtaposed urban area contour polynomials that might coalesce to form megalopolises with high probability could be found. Distance similarity between (difference in number and area of urban sprawl contour polynomials) VIIRS imagery of 2012 and 2016 is a measure of urban expansion, creation of new urban sprawls and merger of adjoining urban areas and suburbs which should be commensurate with rate of growth of GDP between 2012 and 2016. As a space filling problem, this is equivalent to finding distance between 2 temporally separate unequal polynomial (or circle) packings of a 2D surface. As opposed to 4-colored segmented facegraphs of Urban sprawl GIS imagery (Residential,Commercial,Manufacturing-IT-ITES,Greenery) for resource allocation, transportation network graph (e.g retrieved from Google Roads API) could be basis for urban sprawl analytics based on which drones could navigate. Transportation network graph of urban sprawls is topologically and geometrically different from theoretical graphs - edges (rail,roads) between intersections or localities are not necessarily straight lines but are curves of arbitrary shapes, though drone and UAV navigation is mostly along straightlines except obstacle avoidance. In other words, edges (roads) are embedded on a 2D plane (with the exception of multiplanar intersections) which is a product homotopy in topological terms (Analogy: In NeuronRain handwriting recognition by Product Homotopy, Pasting lemma merges paths defined by all possible curve fragments of written alphabets between points x0-x1 and x1-x2). 8 Queens Problem in algorithmic theory which is about dynamic programming based optimal placement of 8 queens on a 64 square chequered chessboard so that none of the paths of queens collide, could be adapted to obstacle avoidance in drone navigation and autonomous vehicles. Queen moves on 8 directions on chessboard cellular automaton grid separated by 45 degrees quite similar to cellular automaton graph model of how pandemics/memes/fads/opinions/cybercrimes spread in social and electronic networks - in order to avoid infection in ER-SIR model, human vertices in a social-urban-sprawl network (which is a floating population) might choose routes found by N-Queen problem (N=average population density, significantly intractable compared to 8 queens). On the contrary, problem of autonomous combat between 2 sets of adversarial drone swarms (https://en.wikipedia.org/wiki/Unmanned_combat_aerial_vehicle) could be theoretically shown to be a PSPACE-complete problem, by an oversimplified reduction from N*N Chess (On the complexity of Chess - N*N Chess is proved to be PSPACE-complete - https://www.sciencedirect.com/science/article/pii/0022000083900302) in which each set (Black and White) has 1 King, atleast 1 Queen (q) and N^h pawns (h > 0) - 2 sets of Adversarial Drone swarms each of size (N^h + q + 1) navigate along N*N 2-dimensional Cellular Automaton Chess Grid. Earlier PSPACE-complete reduction from Chess is too simplistic because actual drone combat is chaotic and drone navigations are along 26 directions atleast on 3-dimensional cellular automaton grid (8 directions on 2D plane + 9-up + 9-down). Almost all Drone Delivery problems are NP-Complete Hamiltonian problems which involve optimal tour mission plans along edges and vertices of an aerial or surface transportation graph. NASA Mars Rover Collaborative Information Portal is an example of hugely successful Robotic Drone Rover which has a scalable streamer service (Service Oriented Architecture) for Mission data transport between Mars Rover File Servers and multiple clients on Earth (Reference: Beautiful Code - Chapter 20 - Higly Reliable Enterprise System for NASA's Mars Rover Mission). UAVs for Drone EVMs, Autonomous Delivery et al., might similarly have to service multithreaded, parallel terrestrial MAVSDK clients which issue imageing commands e.g TAKE_PHOTO. Mars Rover which has been designed for unpredictable terrains and weather could be suitable as platform for online shopping delivery.Online shopping delivery which is still in conceptual stages of R&D has to focus on UAV transport of FMCG, Home appliances and Groceries in the range of 1 kg to 100 kg which are mostly sensitive to handle. Manned drones on the other hand are the holy grail of drone human transport (References on manned drones, flight time (20 minutes) and payload (220 kg): https://en.wikipedia.org/wiki/Passenger_drone, https://lucidcam.com/how-much-weight-can-a-drone-carry/, http://www.onemandrone.com/about, https://www.droneblog.com/how-long-does-a-drone-battery-last-what-you-need-to-know/). NeuronRain theoretical VREPAT (Voter Received Encrypted Paper Audit Trail) voting machine for Drone EVMs envisages an One Time Password facility (preferably non-digital: Currency OTP) and Voter Receipt of an Encrypted Audit Trail for ensuring free,fair and unique Vote - Drone VREPAT EVM tightens and improves upon polling station VVPAT EVMs by : (*) obviating necessity for polling booths (*) Geographic and Demographic uniqueness through aerially navigating to voter's residence thus putting to rest any speculation on impersonation and malpractice at polling booths (*) authenticating the voter non-digitally by currency UUID OTP which is nationwide unique at any point in time - only one person in the country can hold a paper currency of a UUID at any time discounting counterfeits (*) by encrypting voter's choice (RSA or MD5 hash which make voting as secure,secretive and discreet as e-commerce) in a receipt issued to voter which can be decrypted only by an authority in case of legal dispute - EVM malfunctioning and tampering is done away with. Security Analysis of Diebold AccuVoteTS voting machine - [Ariel J. Feldman*, J. Alex Halderman*, and Edward W. Felten*,†] - https://www.usenix.org/legacy/event/evt07/tech/full_papers/feldman/feldman_html/index.html and source code review of Diebold - https://votingsystems.cdn.sos.ca.gov/oversight/ttbr/diebold-source-public-jul29.pdf - finds serious vulnerabilities in DRE (Direct Recording Electronic) machines - any EVM storing votes in memory is prone to attacks necessitating a non-digital complement of every vote. 
+Segregating Non-residential as Commercial or Central Business District (CBD) and Industrial (Manufacturing-IT-ITES) splits earlier 3-color classification into desired 4-coloring and there are kernel density based clustering solutions suggested for identifying CBD : 1) Building an Urban Spatial Structure from Urban Land Use Data: An Example Using Automated Recognition of the City Centre - https://www.mdpi.com/2220-9964/6/4/122/htm 2) Delimitating Urban Commercial Central Districts by Combining Kernel Density Estimation and Road Intersections: A Case Study in Nanjing City, China - https://www.mdpi.com/2220-9964/8/2/93/htm . Moran's I and Gini Index are measures of dispersion and unequal lopsided distribution of growth within an urban sprawl. Graph connecting nightlights circular vertices by edges, could be an aerial transportation graph for autonomous delivery Drones between cities. Electricity Consumption and NightLights are mostly limited to Residential,Commercial and Manufacturing-IT-ITES regions of an urban sprawl and Agricultural land-Greenery-Waterbodies are excluded because of almost nil night lighting. Thus only 3 of the 4 colors of an Urban sprawl FaceGraph are covered by NightLights urban sprawl estimation excluding Farm lands which could be an error-free guideline for Urban planning and Expansion. There is an unusual theoretical aspect to NASA VIIRS NightLights urban area circles - Circular Urban areas of varied radii are packed (close or loose) on 2 dimensional space of a nation which makes it a variant of Space filling or Unequal Circle packing problem. Some bustling urban areas in VIIRS imagery of India 2017 are not heavily lit which could be inferred as "NightLights imply urbanization while converse may not be true (or) NightLights sprawl lowerbounds an Urban sprawl". NeuronRain implements ranking of urban sprawls based on areas of segmented Contour polynomials. Circles or contour polynomials bounding urban areas unequally fill the 2D space and ratio of sum of urban contour areas to area of the country is the urbanization percentage metric. Set of Urban area contours could be grouped by unsupervised clustering algorithms (e.g Neural Networks, DBSCAN, Voronoi Tessellation) using euclidean distances amongst their centroids from which juxtaposed urban area contour polynomials that might coalesce to form megalopolises with high probability could be found. Distance similarity between (difference in number and area of urban sprawl contour polynomials) VIIRS imagery of 2012 and 2016 is a measure of urban expansion, creation of new urban sprawls and merger of adjoining urban areas and suburbs which should be commensurate with rate of growth of GDP between 2012 and 2016. As a space filling problem, this is equivalent to finding distance between 2 temporally separate unequal polynomial (or circle) packings of a 2D surface. As opposed to 4-colored segmented facegraphs of Urban sprawl GIS imagery (Residential,Commercial,Manufacturing-IT-ITES,Greenery) for resource allocation, transportation network graph (e.g retrieved from Google Roads API) could be basis for urban sprawl analytics based on which drones could navigate. Transportation network graph of urban sprawls is topologically and geometrically different from theoretical graphs - edges (rail,roads) between intersections or localities are not necessarily straight lines but are curves of arbitrary shapes, though drone and UAV navigation is mostly along straightlines except obstacle avoidance. In other words, edges (roads) are embedded on a 2D plane (with the exception of multiplanar intersections) which is a product homotopy in topological terms (Analogy: In NeuronRain handwriting recognition by Product Homotopy, Pasting lemma merges paths defined by all possible curve fragments of written alphabets between points x0-x1 and x1-x2). 8 Queens Problem in algorithmic theory which is about dynamic programming based optimal placement of 8 queens on a 64 square chequered chessboard so that none of the paths of queens collide, could be adapted to obstacle avoidance in drone navigation and autonomous vehicles. Queen moves on 8 directions on chessboard cellular automaton grid separated by 45 degrees quite similar to cellular automaton graph model of how pandemics/memes/fads/opinions/cybercrimes spread in social and electronic networks - in order to avoid infection in ER-SIR model, human vertices in a social-urban-sprawl network (which is a floating population) might choose routes found by N-Queen problem (N=average population density, significantly intractable compared to 8 queens). On the contrary, problem of autonomous combat between 2 sets of adversarial drone swarms (https://en.wikipedia.org/wiki/Unmanned_combat_aerial_vehicle) could be theoretically shown to be a PSPACE-complete problem, by an oversimplified reduction from N*N Chess (On the complexity of Chess - N*N Chess is proved to be PSPACE-complete - https://www.sciencedirect.com/science/article/pii/0022000083900302) in which each set (Black and White) has 1 King, atleast 1 Queen (q) and N^h pawns (h > 0) - 2 sets of Adversarial Drone swarms each of size (N^h + q + 1) navigate along N*N 2-dimensional Cellular Automaton Chess Grid. Earlier PSPACE-complete reduction from Chess is too simplistic because actual drone combat is chaotic and drone navigations are along 26 directions atleast on 3-dimensional cellular automaton grid (8 directions on 2D plane + 9-up + 9-down). Almost all Drone Delivery problems are NP-Complete Hamiltonian problems which involve optimal tour mission plans along edges and vertices of an aerial or surface transportation graph. NASA Mars Rover Collaborative Information Portal is an example of hugely successful Robotic Drone Rover which has a scalable streamer service (Service Oriented Architecture) for Mission data transport between Mars Rover File Servers and multiple clients on Earth (Reference: Beautiful Code - Chapter 20 - Higly Reliable Enterprise System for NASA's Mars Rover Mission). UAVs for Drone EVMs, Autonomous Delivery et al., might similarly have to service multithreaded, parallel terrestrial MAVSDK clients which issue imageing commands e.g TAKE_PHOTO. Mars Rover which has been designed for unpredictable terrains and weather could be suitable as platform for online shopping delivery.Online shopping delivery which is still in conceptual stages of R&D has to focus on UAV transport of FMCG, Home appliances and Groceries in the range of 1 kg to 100 kg which are mostly sensitive to handle. Manned drones on the other hand are the holy grail of drone human transport (References on manned drones, flight time (20 minutes) and payload (220 kg): https://en.wikipedia.org/wiki/Passenger_drone, https://lucidcam.com/how-much-weight-can-a-drone-carry/, http://www.onemandrone.com/about, https://www.droneblog.com/how-long-does-a-drone-battery-last-what-you-need-to-know/). NeuronRain theoretical VREPAT (Voter Received Encrypted Paper Audit Trail) voting machine for Drone EVMs envisages an One Time Password facility (preferably non-digital: Currency OTP) and Voter Receipt of an Encrypted Audit Trail for ensuring free,fair and unique Vote - Drone VREPAT EVM tightens and improves upon polling station VVPAT EVMs by : (*) obviating necessity for polling booths (*) Geographic and Demographic uniqueness through aerially navigating to voter's residence thus putting to rest any speculation on impersonation and malpractice at polling booths (*) authenticating the voter non-digitally by currency UUID OTP which is nationwide unique at any point in time - only one person in the country can hold a paper currency of a UUID at any time discounting counterfeits (*) by encrypting voter's choice (RSA or MD5 hash which make voting as secure,secretive and discreet as e-commerce) in a receipt issued to voter which can be decrypted only by an authority in case of legal dispute - EVM malfunctioning and tampering is done away with. Security Analysis of Diebold AccuVoteTS voting machine - [Ariel J. Feldman*, J. Alex Halderman*, and Edward W. Felten*,†] - https://www.usenix.org/legacy/event/evt07/tech/full_papers/feldman/feldman_html/index.html and source code review of Diebold - https://votingsystems.cdn.sos.ca.gov/oversight/ttbr/diebold-source-public-jul29.pdf - finds serious vulnerabilities in DRE (Direct Recording Electronic) machines - any EVM storing votes in memory is prone to attacks necessitating a non-digital complement of every vote. VREPAT Electronic Voting Machines are theoretically based on Set Partitions and fit into categories of Receipt based cryptographic EVMs discussed in paper - "On the notion of “software-independence” in voting systems" - [Rivest] - https://people.csail.mit.edu/rivest/RivestWack-OnTheNotionOfSoftwareIndependenceInVotingSystems.pdf - "... Receipt-based cryptographic voting systems involve a physical, e.g., paper receipt that the voter can use to verify, during the process of voting, whether his or her ballot was captured correctly. The contents of the receipt, in general, employ cryptography in some form so that the voter is able to verify that the votes were recorded accurately; the receipt does not show how the voter voted...." - a value addition in VREPAT is incorporation of Drones to ensure voter's residential address and most importantly physical presence of voter at the address during voting. Recent programming languages like Rust enforce move semantics of variables in memory by default and rule out copies which is a desired behaviour in cryptocurrencies and voting - DRE Voting machines and Cryptocurrencies written in Rust could eliminate replication of votes and currency UUIDs respectively (NeuronRain Boost C++ implementation of Neuro cryptocurrency is std::move() based which is not a default behavior in C++). 
 
 **Can NeuronRain be deployed on Mobile processors?**