vault backup: 2024-10-21 00:46:21

Autonomous Networking/notes/5 Drones.md

@@ -0,0 +1,138 @@
+#### Unmanned Aerial Vehicle (UAV)
+- UAV, commonly known as a Drone, is an aircraft without a human pilot aboard (unmanned or uncrewed)
+- it operates
+	- under remote control by a human
+	- autonomously by on board computers
+
+**Weight:** from 0.5g 15000kg
+**Maximum speed:** up to 11265Kph
+**Propellant:**
+- fossil fuel
+- battery
+
+#### Usages
+- can provide timely disaster warnings
+- medical supplies
+- dangerous situations
+- traffic monitoring, wind estimation, remote sensing...
+
+I many scenarios, UAVs need to exchange a relatively large amount of data among themselves and/or a control station. In many case there isn't any network infrastructure available. Drones can also used to expand terrestrial communication networks.
+
+Drones can be equipped with several standard radio modules:
+- Wi-Fi
+- Cellular
+- LPWAN (low power wide area network, es. LoRa)
+
+## Routing
+may require multi-hop data connections
+#### Comparison WSN with Dronets
+
+|                    | **WSN**                           | **Dronet**                       |
+| ------------------ | --------------------------------- | -------------------------------- |
+| **Mobility**       | none or partial                   | high, even 3D                    |
+| **Topology**       | random, star, ad-hoc node failure | mesh                             |
+| **Infrastructure** | absent                            | absent                           |
+| **Energy source**  | battery                           | battery (very limited)           |
+| **Typical use**    | environmental monitoring          | rescue, monitoring, surveillance |
+
+**Goals of routing protocols:**
+- increase delivery ratio
+- loop freedom
+- low overhead
+- reduce delays
+- energy consumption
+- scalability
+
+### Proactive routing protocols
+Are they suitable for UAV networks?
+- slow reaction to topology changes, will cause delays
+- bandwidth constraints
+
+Protocols:
+- OLSR - Optimize Link State Routing
+- DSDV - Destination-Sequenced Distance Vector
+- B.A.T.M.A.N. - Better Approach to Mobile Ad Hoc Network
+
+### Reactive protocols
+- DSR - Dynamic Source Routing
+- AODV - Ad hoc On Demand Distance Vector
+
+#### Hybrid protocols
+- ZRP - Zone Routing Protocol
+- TORA - Temporarily Ordered Routing Algorithm
+
+### B.A.T.M.A.N.
+A proactive, distance-vector routing protocol for Mobile Ad-hoc Networks (MANETs) and Mesh Networks. Designed for decentralized decision-making and self-organizing networks.
+
+**Key features:**
+- decentralized routing
+	- no node has global knowledge of the entire network
+- next-hop based
+	- nodes only know their best-hop neighbor for reaching a destination
+- link quality driven
+	- decisions are based on the quality of the link between nodes
+- self-healing
+	- adapts to changes automatically
+
+**How batman works**
+- Originator messages (OGMs):
+	- broadcast to announce its presence
+	- OGMs are forwarded by neighbors to propagate through the network
+	- each OGM carries a sequence number to ensure the information is up-to-date and avoid routing loops
+	- nodes evaluates how frequently they receive OGMs to their neighbors to determine link quality
+	- each node maintains a routing table with the best next-hop neighbor based on link quality
+
+- fields inside OGM
+	- originator address
+	- sequence number
+	- TTL (hop limit)
+	- LQ (quality between the sender and the originator)
+	- hop count
+
+Asimmetry problem:
+If A can reach B well, B thinks it can reach A well too. But it may not be the case.
+To overcome the issue there is the Transmit Quality (TQ) algorithm.
+
+B transmits RQ (receive quality) packet to A. A counts them to know the link quality.
+A knows the echo quality by counting the rebroadcasts of its own OGMs from its neighbors.
+Dividing echo quality by receiving quality, A can calculate the transmit quality.
+
+**propagation**
+A when originates the OGMs, it sets TQ to 100%. The neighbor computes their own local link quality into the received TQ value and rebroadcast the packet.
+- $TQ = TQ_{incoming} * TQ_{local}$
+
+![[Pasted image 20241017154152.png]]
+
+### Geographic protocols
+the geographical position information of the nodes is utilized for forwarding decisions.
+Nodes knows their position by GPS.
+
+Geographic routing schemes don't need the entire network information
+- no routing discovery
+- no routing tables
+- forward packet based on local information
+- less overhead, bandwidth and so energy consumption
+- for routing decisions a drone needs only the neighbors and destination position
+
+Every node has coordinates of neighbors.
+
+**Dead end problem**
+several techniques have been defined in sensor networks to recover from a dead end but they are often not applicable to dronets.
+
+Geo routing is based on three main approaches
+
+##### Greedy forwarding
+as stated before
+
+#### Store-carry and forward
+When the network is intermittently connected, forwarder nodes do not have any a solution to find a relay node. Not possible to forward any data packet to a predefined node which is not in range. So the current node will carry the packet until it meet another node or the destination target itself.
+
+#### Prediction
+based on geographical location, direction, and speed to predict the future position of a given node. They will predict the position of a next relay node.
+
+### DGA algorithm
+![[Pasted image 20241017161724.png]]
+![[Pasted image 20241017161747.png]]
+
+![[Pasted image 20241017161803.png]]
+

Autonomous Networking/notes/6 Internet of Things.md

@@ -0,0 +1,55 @@
+IoT term is used to refer to
+- the resulting globlal network of connecting smart objects
+- the protocols ...
+- ...
+
+required features:
+- devices hetereogeneity
+- scalability
+- data ubiquitous data exchange
+- energy-optimized solutions
+
+
+#### Backscattering
+- allows devices to run without battery
+- only available at research level for now
+- use radio frequency signals as power source
+- two types
+	- ambient
+	- rfid
+
+##### Ambient backscattering
+- devices harvest power from signals available in the environment
+- they use existing RF signals without requiring any additional
+
+- Performance drawbacks
+	- low data rate (about 1kbps)
+		- not suitable for real-time applications that continuously exchange data
+	- availability of signals
+		- signal may not be available indoor or not powerful enough
+
+##### RFID backscattering
+
+...
+
+##### Battery free smart home
+- in a smart home there may be a lot of smart devices
+- if every one of them has a battery, it's not good for the environment
+- we can deploy an RFID reader with multiple antennas that covers all the different rooms
+
+### Communication
+add scheme slide
+
+RFID tags run EPC Global Standard
+- in a smart home we may want less bits dedicated to the tag id and more dedicated to the actual data
+- 8 bits for ID
+- 6 bits for data
+- 4 for CRC
+
+### Infrastructure based wireless networks
+- base stations connected to wired backbone network
+- stations choses the closest base station
+- Limits
+	- when no infrastructure is available
+	- expensive/inconvenient to setup
+	- when there is no time to set it up

Autonomous Networking/notes/6.1 RL.md

@@ -0,0 +1,122 @@
+Case study: battery-free smart home
+- each device produces a new data sample with a rate that depends on the environment and the user (continuously, event based / on demand...)
+- a device should only transmit when it has new data
+	- but in backscattering-based networks they need to be queried by the receiver
+
+In which order should the reader query tags?
+- assume prefixed timeslots
+- TDMA with random access performs poorly
+- TDMA with fixed assignment also does (wasted queries)
+- we want to query devices that have new data samples and avoid
+	- data loss
+	- redundant queries
+
+Goal: design a mac protocol that adapts to all of this.
+One possibility is to use Reinforcement Learning
+
+#### Reinforcement learning
+How can an intelligent agent learns to make a good sequence of decisions
+
+- an agent can figure out how the world works by trying things and see what happens
+- is what people and animals do
+- we explore a computational approach to learning from interaction
+	- goal-directed learning from interaction
+
+RL is learning what to do, it presents two main characteristics:
+- trial and error search
+- delayed reward
+
+- sensation, action and goal are the 3 main aspects of a reinforcement learning method
+- a learning agents must be able to
+	- sense the state of the environment
+	- take actions that affects the state
+
+Difference from other ML
+- no supervisor
+- feedback may be delayed
+- time matters
+- agent action affects future decisions
+- ...
+- online learning
+
+Learning online
+- learning while interacting with an ever changing world
+- we expect agents to get things wrong, to refine their understanding as they go
+- the world is not static, agents continuously encounter new situations
+
+RL applications:
+- self driving cars
+- engineering
+- healthcare
+- news recommendation
+- ...
+
+Rewards
+- a reward is a scalar feedback signal (a number)
+- reward Rt indicates how well the agent is doing at step t
+- the agent should maximize cumulative reward
+
+RL based on the reward hypotesis
+all goals can be described by the maximization of expected cumulative rewards
+
+communication in battery free environments
+- positive rewards if the queried device has new data
+- else negative
+
+Challenge:
+- tradeoff between exploration and exploitation
+- to obtain a lot of reward a RL agent must prefer action that it tried and ...
+- ...
+
+exploration vs exploitation dilemma:
+- comes from incomplete information: we need to gather enough information to make best overall decisions while keeping the risk under control
+- exploitation: we take advanced of the best option we know
+- exploration: test new decisions
+
+### A general RL framework
+at each timestamp the agent
+- executes action
+- receives observation
+- receives scalar reward
+
+the environment
+...
+
+agent state: the view of the agent on the environment state, is a function of history
+- the function of the history is involved in taking the next decision
+- the state representation defines what happens next
+- ...
+
+#### Inside the agent
+one or more of these components
+- **Policy:** agent's behavior function
+	- defines what to do (behavior at a given time)
+	- maps state to action
+	- core of the RL agent
+	- the policy is altered based on the reward
+	- may be
+		- deterministic: single function of the state
+		- stochastic: specifying probabilities for each actions
+			- reward changes probabilities
+- **Value function:**
+	- specifies what's good in the long run
+	- is a prediction of future reward
+	- used to evaluate the goodness/badness of states
+	- values are prediction of rewards
+	- Vp(s) = Ep[yRt+1 + y^2Rt+2 ... | St = s]
+- **Model:**
+	- predicts what the environment will do next
+	- many problems are model free
+
+back to the original problem:
+- n devices
+- each devices produces new data with rate_i
+- in which order should the reader query tags?
+- formulate as an RL problem
+	- agent is the reder
+	- one action per device (query)
+	- rewards:
+		- positive when querying a device with new data
+		- negative if it has no data
+		- what to do if the device has lost data?
+	- state?