Problem Definition (Optimal Temporal Paths)

A temporal edge is $(u,v,t,\lambda)$ meaning edge $(u,v)$ is usable at time $t$ and traversal takes $\lambda\ge 0$ (in point models set $\lambda\in{0,1}$).

A temporal path / journey $P=e_1,\dots,e_k$ with $e_i=(v_{i-1},v_i,t_i,\lambda_i)$ is time-respecting if

We write:

By default, waiting at intermediate vertices is allowed. A release time $t_0$ restricts attention to journeys with $\mathrm{dep}(P)\ge t_0$.

📁 Definition: Earliest-Arrival (Foremost) Path

Given $s,t\in V$ and $t_0$, an earliest-arrival (aka foremost) path minimises $\mathrm{arr}(P)$ among all journeys $P:s\rightsquigarrow t$ with $\mathrm{dep}(P)\ge t_0$.

📁 Definition: Shortest (Fewest-Hops) Temporal Path

Given $s,t,t_0$, a shortest temporal path minimises the hop count $|P|$ among all feasible journeys. (Common tie-breakers: smaller $\mathrm{arr}(P)$ or smaller $\mathrm{dur}(P)$.)

📁 Definition: Fastest (Minimum-Duration) Temporal Path

Given $s,t,t_0$, a fastest temporal path minimises $\mathrm{dur}(P)$ among all feasible journeys.


Design Notes & Relationships